`generic_arg_infer` impl is prone to bugs in hir visitors

[BoxyUwU]

The current issue with the implementation is that there are too many ways of representing infer vars in the hir. We have TyKind::Infer, GenericArg::Infer, and ArrayLen::Infer. It would be too easy for someone to check only for a TyKind::Infer generic arg but forget about GenericArg::Infer, or check both of those but forget about ArrayLen::Infer.

At this moment I'm not entirely sure what the best way of fixing this is. rust-lang/compiler-team#480 was an attempt to fix this but it was too broad (attempted to also support unbraced const param arguments), did not solve ArrayLen::Infer, and I'm not even sure how nice removing TyKind::Infer would be since we use hir::Ty in places for things that arent generic args.

Originally posted by @BoxyUwU in #85077 (comment)

[fogti]

but that other issue seems to indicate that the problem is also somewhat based upon ambiguities in the language itself, and it seems somewhat that the implementation just passes along this ambiguity (which I would also do if I were confronted with this problem). so it might be the case (I don't have enough overview about RFCs and specific implementation to know for sure) that this would need some deeper changes to be fixed (maybe including changing unstable syntax), could also indicate that maybe? an RFC might be needed to resolve this?