Okay, with a full rebuttal to your Syncytin argument.
I chose not to make this a full technical dissertation, but i have a collection of papers, diagrams, and all the gritty details to back it up. Just, please, have mercy. Please don't punish me for this by making me into your dancing monkey running around as your research assistant. 🙏

"ERV's are real, they are insertions of viral DNA in infected cells."
That's right, they make up 8% of our genome, and we can identify them by their clear retroviral signatures, e.g.,
[5' LTR]-[~2 kb insert/intron]-[gag pseudogene]-[pol pseudogene]-[env ORF = Syncytin-1]-[3' LTR]

"Fixation of ERV genetic info into a host species is extraordinarily rare"
Yes, and it must be a lot rarer than zygote infection by an ERV.

"Germline infection is rare"
Not as much as you might think.

This only requires the infection to be upstream of the formation of a zygote for it to end up in a zygote, like it would within any other cell lineage.

8% of our genome is made up of ERVs that managed not to kill their host, managed to survive the random sorting of gene flow and genetic drift, and were lucky enough to find their way into a lineage that went on to become ancestors to all present-day humans. For any one of these, there are legions more that infected a zygote but didn't survive these odds.
Keep that point in mind.

Now, can you think of something that could help an ERV survive these odds?
That's right, conferring a benefit to its host.

Positive selection amplifies the rate at which an allele will reach fixation. That means that beneficial ERV insertions would be overrepresented among those that did manage to beat the odds and reach fixation.

But, conferring a benefit in this case is a tall order, though, right?
Not as much as you might think.

Unlike any old mutation, the ENV gene is already a premade snippet of working code. And, being from a virus, it comes prepackaged with an immunosuppressive domain that safely evades the host's immune system. The Syncytin gene really is just that, premade ENV code for producing ENVelope glycoproteins.

Problem: The placenta needs to fuse cells.
Solution: The Virus already has a fusion engine (Env) to enter cells.
Problem: The placenta needs to hide from the mother's immune system.
Solution: The Virus already has an Immunosuppressive Domain (ISD) to evade the host's immune system.

So, all it needs to do is safely land in the right spot.
Tall order, right?
Not as much as you might think.

Unlike prokaryotes, which, given their large population size, can afford the cost of having the kind of tightly conserved string of code that you imagine all genomes to be, eukaryotic genomes are almost entirely nonconserved and nonconstrainted genetic material. There is a lot of safe intergenic space. The ERV does not have to land smack inside a "placental gene". It only needs to land safely within a neighbouring intron.

Doing this 9 times is a tall order, though, right?

Well, maybe, but this isn't like landing on the head of a pin. Those legions of ERVs have a target space much wider than you think. These 9 Syncytin cases did not all land in the same spot.
These are 9 different genes making use of envelope glycoproteins, not the same gene copied over into 9 different mammal groups.
Whatever the odds, only 9 individuals out of all mammals throughout history had to be so lucky.

Many mammal groups did not inherit this placental upgrade, and they get by just fine, so it's also not the case that this needed to happen at all. The Opossum doesn't even have a placenta.

Now, whatever you think odds might be, given all this in mind, the odds that any two ERV insertions would actually land in the exact same spot in two individuals, with both of these reaching fixation while not conferring a benefit to their hosts, would be many orders of magnitude smaller. It's not impossible, but much less likely than those 9 Syncytin examples.

We're well past bananas.

Now compound those odds with the odds that these two individuals would just so happen to fall in the right order, one within the subclade of the other.

Now repeat those crazy odds all the way down the phylogenetic tree. This needs to happen over and over again on every other side of each kind barrier from the first clade through all subclades up to the present day.

I want to really drive this point home.
This compounding set of EXTRAORDINARILY unlikely events needs to be repeated in a pattern that just so happens to follow the same phylogenetic path as every other signal of common descent that you need an independent explanation to handwave away.

And were not even done yet.
You now have to compound these extra-extraordinarily unlikely odds over again for every single real example of a shared ERV insertion. That 8% of our genome minus the human-exclusive.

And like i've said before, we can repeat this same matching phylogenetic signal argument, with similar kinds of odds, for, transposable elements, nuclear mitochondrial DNA insertions, chromosome fusion events and rearrangements with matching breakpoints, mitochondrial and chloroplast gene order rearrangements, intron position gains/losses, conserved “signature” indels, shared pseudogenization events, gene duplication events with shared duplication boundaries, microRNA family presence/absence, all the silent site variations on spelling for the same genes, AND the decay signals within and around each of these.
Just to name the major ones.

Now some of these you might think are much easier to explain away than others. They are not all the same kind of event. They each have their own particulars. But, you would have to come up with independent arguments for the particulars of each, with at least the same force as your Syncytins objection, over again, independently for every single one of these.
Otherwise, the odds keep on compounding.

So, what's more likely?
Do all these signals match by chance, by design, or could it possibly be that the signal and corresponding noise and pattern of signal decay have been shaped by an actual line of inheritance down an actual family tree?

I don't expect you to fully accept that, but maybe you might now see why some of us find it reasonable to conclude that it just might be inheritance that explains this pattern.
You know, inheritance, the only way we've ever observed genes finding themselves anywhere?

Does it all sound like the same bananas to you now? A small tangent:
"I should be clear "modern" humans. HERV-K (HML-2) insertions are inferred to be older than modern humans, predate the split between humans and Neanderthals some humans and Denisovans"

Wouldn't Neanderthals and Denisovans also be descendants of Adam and Eve?
Wouldn't that make them "modern humans", and the window for ERV fixation rather short?
Either we were created with the HERV-K106 in place, or it would have to have been fixed at least at the flood bottleneck 44 centuries ago, and all polymorphic ERVs would have to have reached the frequencies they are at today sometime down the line from there.

You are missing my point entirely,
Your “40M mutational differences” number is a base-pair difference count, not an event count. Duplications/indels/repeats can create thousands of base differences from one event.

So how do you get the 45 mutations per generation number from?
Where do you get the idea that they would all have to be beneficial?
Are you simply making that up so you can apply Haldane’s cost?

And let's see you apply this same math to Neanderthal and Denisovan divergence in the extremely short period between the Tower of Babel and their respective diasporas across Europe and Asia. While i wait for your response, i thought i'd go over some examples of human-specific mutations that contribute to chimp-human divergence.
Remember, you are not allowed to recognize these as being mutations at all, or even "Preprogrammed Adaptive Variation."

NOTCH2NL. This mutation is a partial duplication of NOTCH2. It dampens the NOTCH2 signal that triggers cells to stop dividing and become neurons. The result being a higher neuronal density in the human brain.

It occurred within the 1q21.1 locus, which is a mutation hotspot, we didn't have to wait around long for recurrent reciprocal deletions and duplications in this region. They happen frequently, causing neural developmental issues.

But, it's not like we were waiting around for it to happen. It just so happened that some population from our ancestors were in a position to benefit from higher neuronal density, and interbred with populations that contributed other human-specific mutations, like the following...

@SDAFS56788141 @LatFilosof @SimonSaysJ15 @KingOfSalt4 @IanCopeland5 18% figure comes from counting the 238 million base pairs identified as "non-syntenic."
By that metric, human to human differences can be up to 6-8%. @SDAFS56788141 @LatFilosof @SimonSaysJ15 @KingOfSalt4 @IanCopeland5 By that metric chimp-gorilla genomes differ by 17.9% to 27.3%.

@SDAFS56788141 @DivinelyDesined Alright. ERVs it is! Okay. So i spent some time thinking about how to address the Argument for Common Descent from ERV Insertion Site Distribution with you, @SDAFS56788141.
I think we could have a productive discussion. I will try to make a case that lends credence to one of two conclusions, using an Argument for Common Descent from ERV Insertion Site Distribution:

1) Universal Metazoan Common Descent is the case,
or,
2) The diversity of Metazoan life was deliberately designed. AND the designer deliberately chose to design Metazoan life such as to make it look like Universal Metazoan Common Descent is the case.

While i make the case for #1, each
"That just so happens to be how the designer chose to do it"
rebuttal will lend further credence to hypothesis #2.

First, let's see if we can set some groundwork.
I made a list of points i'd prefer to take for granted for the sake of the argument. One, it can give you a sense of the general framework i'm starting with, secondly, the more of this we can agree on in at least a general sense, the less of a mess we will have to make of this thread. We can squabble over exact numbers or frequencies, but the general ratios and proportions, i think are fair.

1) Mutations happen. Beneficial, neutral, and deleterious. The fitness effect of a mutation is dictated by selection pressures. A mutation does not have to add new information or create whole new structures to have a net-benefit, that is, to have an above neutral frequency of finding itself in future generations.

2) Observations and simple calculations of Mutation and Drift Load dictate that a large portion of metazoan genomes must be non-conserved and uncontrained, effectively serving as a buffer for the mutation load.

3) Most metazoan mutations happen in the much larger non-conserved and uncontrained regions of the genome, where they are effectively neutral.

4) Much fewer mutations happen within the tiny portion of functionally conserved and constrained regions, but are highly prone to removal or neutralization under strong purifying selection.

5) Beneficial mutations are extremely rare, but highly prone to fixation by positive selection pressures.

6) Beneficial de novo mutations are the rarest of the rare mutations. Mutations are far more likely to be beneficial when using preexisting genetic material within less-conserved, less-constrained regions, such as previous duplications, neutralized EVRs, or pseudogenes.

7) Endogenous Retrovirus insertions happen. ERV germline insertion is rare, stochastic, effectively irreversible, but common enough for us to find them fixed across metazoan genomes.

8) Given survivorship sampling bias -seeing only what managed to get fixed through Drift or positive selection- we should see a highly overrepresented sample of beneficial, neutral, or neutralized mutations.

9) If EVR exaptation ever did happen, the same survivorship sampling bias would apply.

10) There are several Retroviral functions that have potential exaptation value.
For example,

10a) The Env Gene: Using the "Jamming the Lock" Antiviral Defense. The polymorphic EVR exaptation "Friend Virus Susceptibility 4" (FV4) provides immunity to leukemia by expressing a "broken" copy of the viral Env protein that blocks entry to particular cell receptors.

10b) The LTRs (Long Terminal Repeats): Viral promoters are ready-made "switches" for tissue-specific host gene regulation. Dispersed insertion sites could potentially coordinate simultaneous expression across complex genetic networks.
10c) The Pol Gene: Viral "cut and paste" integrase machinery exapted for the somatic recombination necessary for adaptive immunity (RAG1/RAG2).
10d) The Gag Gene: Viral capsid packaging for neuronal RNA transport (Arc), capsid assembly proteins as "decoys" to disrupt infectious viral assembly (Fv1).
10e) Or simply "Junk" Spacer Material: Non-coding viral spacers for repetitive protein structures, such as antifreeze glycoproteins.
11) If any two kinds were independently created, they should have independent histories of mutation accumulation within non-conserved and uncontrained regions.

Below is an outline of the Common Descent Argument via ERV Distribution in Apes/Humans to work with.

Premises
P1) ERV Insertion Site Specificity:
An Endogenous Retrovirus (ERV) insertion into a germline cell's DNA is a rare, stochastic (random), and effectively irreversible event. Therefore, the initial genomic location (the insertion site) of a new ERV copy is unique for that specific insertion.

P2) Shared Insertion Site Identity Implies Inheritance:
If two or more ape species possess an ERV at the exact same genomic insertion site, the probability that this identity arose from two independent insertion events is astronomically low (due to Premise 1).
Consequently, the shared presence of an ERV at the same site must be due to inheritance from a single ancestral organism in which the original insertion event occurred.

P3) Nested Subset Distribution Pattern:
The distribution of all shared ERV insertions among extant ape species and humans is not random, but forms a nested hierarchical pattern (a nested subset structure). This means:
A large subset of ERV insertions is shared among all apes (e.g., humans, chimpanzees, gorillas, orangutans, gibbons).
A smaller, but still large, subset is shared by a more restricted group (e.g., humans, chimpanzees, and gorillas).
An even smaller subset is shared only between the most closely related pair (e.g., humans and chimpanzees).
The total set of ERV insertions in any one species (e.g., humans) includes all the insertions from the various shared subsets, plus a small, species-specific set of recent insertions.

Conclusion:
The observed nested subset distribution pattern (Premise 3) of shared ERV insertion sites (Premise 1) among apes (including humans) is precisely and uniquely the pattern predicted by the mechanism of inheritance from a common ancestor (Premise 2). Therefore, all apes (including humans) share a common ancestor, and their evolutionary relationships form the nested family tree hierarchy of common descent.

More formally:
P1: The probability of independent retroviral insertions occurring at the exact same genomic locus in separate lineages is negligible due to the stochastic nature of integration across a vast genome (Statistical Axiom).

P2: Multiple distinct ape species possess ERV insertions at identical orthologous loci (Empirical Premise).

C1: Shared ERV insertions in apes are necessarily derived from a single insertion event in a shared ancestor, rather than independent convergence (Modus Tollens from P1 & P2).

P3: The distribution of these shared ancestral insertions constitutes a mathematically precise nested hierarchy of sets and subsets (Empirical Premise).

C2: Apes evolved via common descent, as branching descent with modification is the unique causal mechanism capable of generating a nested hierarchy of inherited traits (Inference to the Best Explanation from C1 & P3).

When were done, if you like, i can repeat an identical Argument for Common Descent from Retroposon Insertion Site Distribution.

Then we could do 10 additional laps around the argument by swapping out the specifics:
1) NUMTs (nuclear mitochondrial DNA insertions)
2) Other endogenous viral elements (EVEs)
3) Chromosome fusion events
4) Shared chromosomal rearrangements with matching breakpoints
5) Mitochondrial (and chloroplast) gene order rearrangements
6) Intron position gains/losses
7) Conserved “signature” indels
8) Shared pseudogenization events
9) Gene duplication events with shared duplication boundaries
10) microRNA family presence/absence

The adaptive variation through minor tweaks to regulatory networks gets a little uncomfortable too close to home.
We can trace just about everything that makes us human to simple little tweaks to say, developmental timing events.

You will like this one.
The SRGAP2C duplication was a truncated copy that disables the original copy. By jamming the "speed up" signal, synapses stay immature longer. This allows for denser synaptic connectivity—a hallmark of human intelligence.

Synaptic Neoteny: The SRGAP2 Duplication
SRGAP2A, promotes the rapid maturation of dendritic spines, limiting their density. A duplication created a truncated copy, SRGAP2C which encodes a truncated F-BAR domain.

The SRGAP2C protein binds to SRGAP2A, creating an insoluble heterodimer that neutralizes SRGAP2A function.

Phenotypic Outcome: In the absence of active SRGAP2A, dendritic spines mature much more slowly. They remain in a thin, filopodia-like state (highly plastic) for longer. Paradoxically, this delay allows the neuron to accumulate a higher density of spines in adulthood. While a chimp's synaptic development might plateau within a few years, a human's continues into the third decade of life.

Citation A (The Discovery of the Duplication):
Dennis, M. Y., Nuttle, X., Sudmant, P. H., Antonacci, F., ... & Eichler, E. E. (2012). Evolution of Human-Specific Neural SRGAP2 Genes by Incomplete Segmental Duplication. Cell, 149(4), 912-922.

Key Finding: This establishes the "what" and "when." It maps the duplication events (SRGAP2A -> B -> C) and dates the emergence of SRGAP2C to roughly 2.4 million years ago (coinciding with the Homo genus transition).

Citation B (The Functional "Jamming" Mechanism):
Charrier, C., Joshi, K., Coutinho-Budd, J., Kim, J. E., ... & Polleux, F. (2012). Inhibition of SRGAP2 Function by Its Human-Specific Paralogues Induces Neoteny during Spine Maturation. Cell, 149(4), 923-935.

Key Finding: This establishes the "how." It shows that SRGAP2C binds to SRGAP2A (the ancestral gene) and inhibits it. This slows down spine maturation (neoteny), which paradoxically allows for higher density and longer periods of plasticity (learning). The Human Brain "Upgrade": NOTCH2NL

The System: The Notch signaling pathway is a conserved "differentiation engine." It tells stem cells when to stop dividing and turn into neurons.

The Event: A partial duplication of the NOTCH2 gene created a non-functional pseudogene.

The "Inference" Trap: We infer from genomic comparisons (human vs. chimp/gorilla) that a gene conversion event "repaired" this broken pseudogene, creating NOTCH2NL.

The Neofunctionalization: The new NOTCH2NL protein interacts with the original NOTCH2 receptor but—crucially—dampens the signal.

Old Logic: High Signal → Stop dividing → Become Neuron.
New Logic: Dampened Signal → Keep dividing → More Stem Cells → Bigger Brain.

Citation A (The Mechanism & Evolutionary History):
Fiddes, I. T., Lodewijk, G. A., Mooring, M., Bosworth, C. M., ... & Haussler, D. (2018). Human-Specific NOTCH2NL Genes Affect Notch Signaling and Cortical Neurogenesis. Cell, 173(6), 1356-1369.

Key Finding:
This paper details the evolutionary history (the duplication, the pseudogene state, and the gene conversion "repair" ~3-4 MYA). It demonstrates how NOTCH2NL interacts with the Notch receptor to delay differentiation, leading to more neurons.

Citation B (The Functional Consequence):
Suzuki, I. K., Gacquer, D., Van Heurck, R., Kumar, D., ... & Vanderhaeghen, P. (2018). Human-Specific NOTCH2NL Genes Expand Cortical Neurogenesis through Notch Signaling. Cell, 173(6), 1370-1384.

Key Finding:
This paper focuses on the functional aspect. They expressed NOTCH2NL in mouse embryos (which don't naturally have it) and observed an expansion of cortical progenitors—essentially making the mouse brain develop more like a human brain.

"Complex, specified information systems like those found in life are always a result of intelligence."

Humans have bred many breeds of dogs for very specific purposes. The information in the genomes of these various breeds must be specific. Even if God wrote the kernel, he didn't write each specific blueprint for each breed. Rearranging code creates previously unwritten programs.

It couldn't be more obvious to anyone that it doesn't require human breeding to rearrange genetic code to create new varieties from earlier forms. We can see quite clearly how simple changes to Hox coding can create a massive variety of organisms from one basic form by multiplying and modifying body segments.

Take HOX Abd. A "Abdominal A", in a Arthropod segmented body plans, like the diagram of Parhyale hawaiensis' segments below. You can see how they are all modified copies of the same module. T4 and T5 are limbs used for walking, T6, T7, T8 are those same modules in reverse, used for jumping.

Turn off HOX Abd. A, the backward jumping legs develop into forward walking legs, and the A1, A2, A3 pleopod swimming appendages develop into the same uropod anchoring legs from A4-6.

With minor changes to the activation or order of arthropod Hox genes, there is an endless variety of arthropods you can make without having to create any new structures. If you are going to claim "preprogrammed adaptive variation", well, you should take a look at just how much adaptive variation exists out there. Here you can see the path from the Pancrustacea clade, like our parhyale hawaiensis friend above, to the Hexapoda body plan of flies.

The Birth of the Head Organizer (zen → bcd):
A gene originally used for protective egg wrappers (zen) was duplicated and re-wired into a maternal "paint brush" (bicoid), allowing the mother to chemically define the head before the embryo even starts building it.
The Defining Feature (Ubx):
The Ubx protein evolved a molecular "stop sign" for limbs, actively suppressing leg growth in the abdomen to carve the precise six-legged insect archetype out of the multi-legged crustacean ancestor.
The New Segmenter (ftz):
Fushi tarazu shed its old role as a slow identity-specifier to become a fast-acting "pair-rule" gene, rapidly subdividing the embryo into segments like a pre-cutting tool for the insect's accelerated development.