> we can still look at the overall population distribution and say something relevant about heritability since most people have pretty normal genes and environments

So if you had some groups with generally fairly different genes and environments both, this would ascribe any differences just to the genes and throw out the environment aspect?