I think the intuition people want to apply here is that ‘is P=NP’ feels like an instance of a class of questions - ‘is P=x’, and if that is computable, then it

I think the intuition people want to apply here is that ‘is P=NP’ feels like an instance of a class of questions - ‘is P=x’, and if that is computable, then it implies we can attack P=NP by coming up with a way to compute it and feeding in NP. But I think the issue there is the assumption that it even makes sense to define a function over ‘classes of questions’.