This looks rather similar to P2927R1 in the rough domain of problems it is trying to solve. That's a successor paper to the two papers you cite in your own references. Have you considered their approach and how it contrasts with yours?

Also, while I'm not against the core idea I'm not entirely convinced by your motivation. Your sole example is getting RTTI from an exception thrown across a shared library boundary; when there may not be a typeinfo for an arbitrary foreign exception in the first place. Indeed, looking back at your original suggestion thread you were warned that a proposal would need to address this possibility and I don't see that in the paper.