As suggested by @tungmetall and after some good points from @EliLea, a few of the negatives/discussion areas building on my thoughts (bit.ly/2ONNIdl) on unmanned turrets yesterday 
Most of these are a case of theory vs current reality, the latter lagging behind the former. One hopes the next gen of AFV catch up a bit as they should have bespokely designed and integrated designs.
(1) An easy one is size. Whilst the potential for very low profile turrets exist, most to date have been quite dumpy and offer no major silhouette advantage over manned options. (Pic: Unmanned CV90CZr left, manned CV90CZ right)
Russia's Epoch is a really slimline offering (thanks to some deck penetration) and the old concept Falcon turret for Challenger 2 shows MBT can be much smaller too.
However more often they are quite large structures - see here the Samson-2 and RCWS-30 unmanned turrets.
Where turrets are added as an upgrade mid-life they tend to be large and raise centre of gravity significantly. Might do some maths to try and demonstrate the mobility impact (side slope etc) of adding a big weight this high on an 8x8 as is popular of late.
(2) Old school SA is degraded significantly. Cant pop your head out the hatch when you dont have one because youre now deep in the hull. Even when there is one such as Stryker M1128 here its going to be in a poor spot for SA as it wont be at the highest point on the veh anymore
Thats not just combat SA - even just guiding the driver in tight spaces becomes impossible as your view is obsructed on at least one side. Stryker Dragoon reported to require someone to dismount to guide it if the vehicle gets anywhere tight to maneuvre.
(3) Placing your entire reliance for SA and weapon operation on 'the machine' without manual reversion exaggerates the existing risk of an 'SA Kill' - where you weapons, mobility etc may be fine, but your SA has been wiped out and you are an effective kill.
When your only means of viewing the world around you and engaging the enemy relies on various cameras, losing those cameras is a big deal. Its not a new issue, but the reliance on technology exacerbates it significantly.
Aside the obvious SAF and blast/frag threats, some novel solutions are arising to defeat AFV ISTAR/SA systems. MBDA/TDW has shown an EMP missile concept, and a few orgs have looked at rounds that dispense ferrous and other particulate to coat optics, jam turret rings and so on.
(4) Reloading. Legacy AFV are generally manually loaded with 3-7 round clips. Modern AFV like Ajax have larger magazines but usually can be reloaded from within the turret. Unmanned almost always require a reload to be effected outside the vehicle, see here the Puma's magazine
Where legacy/manned turrets allow crew to reload from ready racks or other internal stowage as required under armour, unmanned turret has magazine and thats it. Reload means withdrawal to safe location at rear to exit vehicle and reload, a double loss of time and availability
(5) On much the same note, remote turrets also mean the crew are disconnected from the weapon system in respect of jams, mechanical problems and maintenance.
Whilst modern weapons are very reliable, an issue that may be a simple 30 second clearance or intervention in a manned turret could render an unmanned one inoperable and force a withdrawal to access the turret externally. Here the access to Pumas main gun receiver
(6) Semi-tangential as this thread was about IFV class vehs, but unmanned tank turrets mean less crew (lose loader), which has significant impact on how veh maintenance tasks and procedures are carried out in the field.
A traditional 4-man crew has a tough life keeping heavy metal moving, removing a man makes things more miserable. That said Russia, France etc run 3-man without too much issue, so not critical, just painful for the users.
(7) Cost is final issue for now. Unmanned turret is generally more expensive to procure and to support. More compact is great for reasons already discussed, but also means less space to play with for future upgrades and mods - more cost to integrate and less options to play with
Non-exhaustive and relatively rushed list, but some talking points in the debate to chew on when we're looking at the potential that concepts like Carmel (bit.ly/2Mlj8qo) offer us for the future of AFVs. #AFVaDay /end
