I believe this. The problem is, it's not unique to CS (it might be unique to knowledge work, but CS is far from the only species of knowledge work). So for it to have demographic explanatory power, it would have to account for the intense solitary concentration required for elite research in other STEM fields --- including fields like math, which are very closely related to CS, and the multitude of random STEM fields which have been taken over by computational approaches where the underlying work is really just domain-specialized programming by another name.