The comparison with averages is probably misplaced. Compare the 95th percentiles of males and females in math and science and you might get a better idea of who is likely to make it in engineering. I suspect the answer may be the same, but the off-chance it isn’t does make me question the validity of this argument.

I’ll give another example: studies show women are better/safer drivers on average, but the very best drivers – Formula 1, Nascar etc. – tend to be men.

I will also note that many of the other statistics are meaningless without comparators. How much does male self-esteem drop by 10th grade? Is it possible everyone is (on average) more insecure as they go through their teenage years? What proportion of men with a degree in math or science work in a related field? When men are told “women are better at this task” (which is something this article does with respect to IQ tests and high school grades) how does it effect their performance? etc.

Your point on the statistics being less meaningful (I wouldn’t say meaningLESS) without comparisons is a good one.

However, I disagree that Formula 1/Nascar drivers are necessarily the very best drivers, or that because there are fewer women in those fields, men must be the very best drivers.  It’s always important to remember that correlation does not imply causation.  It could simply be that there are no opportunities for women in those sports (likely), but that women are just as good at driving as men are.  Or it could be that men are more attracted to the competitiveness and danger of those sports than women are, but again, not necessarily that they are better drivers.

Just because a field is dominated by a specific type of person does not necessarily imply that that kind of person is better at that field than other kinds of people.  That’s a logical fallacy.