The author suggests that when these average statistics propagate into life decisions and employment preferences, you end up with an equilibrium with less females in these roles.
You write 'This can be used again to claim women don't belong in those roles' but the author did not use this to claim such a thing! 'I know several women' precisely coincides with what the author wrote. Again, he pointed out multiple times throughout that he was not generalizing but was merely looking at average trends.