This is a universal complaint that can apply to almost any heritability study. And I'm not aware of any situations where a further investigation found it to be

This is a universal complaint that can apply to almost any heritability study. And I'm not aware of any situations where a further investigation found it to be devastating to the original conclusion (mainly because actually covering all the edge cases you're calling out is somewhere between "cost prohibitive" and "impossible"). That makes me think it's just a generic "I don't like this conclusion so I'll criticize it" Joker card that people play, and not a legitimate debunking.