I think there's a fairly common (among mathematicians and theorists) understanding of mathematics wherein its truths are just something-like-definitional-consequences of choices we make when looking at axiom systems or measurement frameworks. ...
... so when someone says "2+2 always equals 4" it's natural to think this is something like saying "The symbols "2", "+", "=", and "4" are always such that "2+2 = 4" is a consequence of their proper application", in which case the natural thing to do is check other frameworks...
... and then it turns out that this is not true, since it turns out there are settings where we use these symbols in something-like-but-not-identical-with their typical use and "2+2 = 4" comes out false therein. Maths nerds then get excited about talking about these frameworks...
... but that background philosophy of mathematics, and that understanding of the question and what is at stake is itself very far from universal (in fact I would guess it's a minority of professional mathematicians) and so comes out looking like just a bizarre reply to others...
... if one thinks of mathematical truths in a more realist bent, not just linguistic artefacts or anything of the sort but rather as real descriptive truths (about nature, or peculiar abstract objects, or something else) then what matters isn't the words but what they point to...
... so they'll say "Obviously if you change what the symbols mean the sentence isn't true but who cares about that!?" - but from the initial perspective, this is rather like saying ""Bachelor" always means "unmarried man"" then dismissing comparative linguistics as irrelevant...
... if you didn't care how people might use the symbols/phonemes involved in "Bachelor" in a different way why would you ask in the first place? I think this background difference of often incoherently felt rather than explicitly developed philosophies explains 2+2 discourse.
