The author doesn't say "all women" to the point that a single counterexample disproves the hypothesis. Not only does the auther qualify statements such as "fewer women" but any such analysis is a statistical study of a group. It says something about a group and similar groups but nothing about a specific individual (in a definitive non-probabilistic sense at least).

It's a bit like arguing that the research into the ill-health effects of smoking are nonsense because your Uncle Jack smoked 3 packs a day for 80 years and died at age 95 when hit by a bus (ie not lung cancer).