The reader brings up a common argument–that our culture links maleness to scientific disciplines, and thereby discourages female involvement. I think there’s something to this, but I suspect it stems from a common phenomenon: When group differences crop up in society (in this case, I would argue, because of a male-female biological gap in science ability), people notice them and form stereotypes. When people act on these stereotypes (such as by assuming their daughters won’t want to go into science), they further solidify the group differences.
Thus, it’s fair to say that culture plays a role, but it’s not fair to say that absent these cultural preconceptions, men and women would be equally represented. To eliminate the unfair perception (that all women are bad at science), you’d have to either (A) get people to stop noticing the accurate perception (that women are less likely to be very good at science) or (B) get people to recognize the difference between a disproportion and a categorical difference. Good luck with either of those.
A few additional comments. One, there is some evidence that stereotypes can hurt women who are good at science. For example, if a man and a woman both have the same science ability, the man is more likely to enter a science field. This is in part due to intellectual differences (if a man and a woman are both very good at science, the woman is more likely to also be very good at verbal reasoning, and so might become a lawyer or whatnot), and in part due to the fact that men tend to like science more, but I don’t think it’s unreasonable to suspect that biases come into play–or to note that such biases are unfair.
Also, when it comes to public policy, unless we can find out to what degree the gap has to do with each biology, interests, and cultural pressure, there’s no way to tell what percentage of scientists “should” be women. Therefore, the focus should be on making girls feel comfortable becoming scientists–not on encouraging them to do so when they’d rather do something else, or on lowering standards to tweak the numbers.