I think there's a middle ground. By acknowledging that people come in all shapes and sizes and mindsets, you can consciously find people who compliment you or your organization well.

I never think that biology should be used to discriminate, since the 90th percentile line of women may be the 85th percentile of men (or whatever skew you want), that doesn't mean that she's not better at whatever task than 85% of men, and hence a good hire. Basically, the macro scale says nothing about the micro scale. But what I wanted to point out was that using macro level statistics and saying "this is inherently wrong" is reading the skew, and not the individuals.

As with everything in life, I'm guessing the real answer is some combination of socialization and biology.

One other ideas has been kicking around in my head, a comp-sci idea even... If you have women being approximately equal members of college (I think they have a slight lead by a few percentages), and you have few women in comp-sci, by the pigeonhole principal, they're the majority in another major. Why isn't there a clamor to get men into those majors? The example of biology was used elsewhere in this thread I think.

Yes, some fields have a lot more women than men. The thing is that men can choose a lot more freely (although it is true there is prejudice against men in some areas, like male nurses), while women are facing a lot of prejudice, sometimes very hardcore prejudice, like death and sexual threats (and you can find a lot of incidents well documented). Men need not worry about such things if they choose to engage in a female dominated field.

The main issue, I think, is not whether women opt more or less for CS, but rather that many many are compelled/driven to opt-out of it. Something is clearly wrong if they can't opt freely without suffering prejudice or without having to, as some say, "grow a thick skin".