I decided to put this one under "Science" because I think it's really a case of people reacting badly to an empirical study. As a statistician, it's true that women routinely (and statistically significantly) score in the mid to upper percentile range on tests, including maths and science tests. It's also true that men score at the very high and very low percentiles. Just because you say this out loud doesn't mean you think that women are in any way better or worse at maths and science. However, as scientists and statisticians we should be allowed to speculate on why this is. Are women genetically worse at maths? Does society lead them to doubt their abilities in the numerical arts? (from personal experience, this is a big YES.) Do women get shunted out of maths and hard sciences early on in high school and then find it impossible to catch up? (Also a YES - story to follow).
I honestly believe Summers was set up for a fall here. Here is my evidence:
--Last year there were far too few tenure offers made to women at Harvard. However, this isn't something that you can just change immediately - you can't send the departments back to the drawing board to ask them to submit more women in a week's time. The tenure process is a long one based on publication, teaching, and general scholarship, and if you want more women to get tenure you take more of them into PhD programs and post-docs and assistant professors. You can't suddenly give undeserving women tenure too early in the hope of redressing your statistics. This hurts both the students, by possibly giving them unworthy professors, and the professors themselves by leaving them wide open to accusations of insufficient scholarship and preparation. To emphasize again - the low number of women tenured ISN'T Summers' fault. (In fact, since he doesn't pick the professors who get tenure at all, if it's anyone's fault, it's the departmental chairs.)
--Summers also has a history of butchering his words, although this doesn't excuse any potential gaffes he may make.
--Knowing this, someone asked Summers to speak at this conference about women in academia. According to the Crimson, Summers was assured:
a) that it was a scientific conference
b) that it would be completely off the record precisely because of worries that comments would be misconstrued
c) that he would be there in his capacity as an economist, and not speaking as the Harvardian mouthpiece.
What actually happened was:
a) he summarized the contents of a paper that was about to be presented at the conference, which had valid statistical findings with regards to women underperforming in the sciences
b) his comments were reported in the Boston Globe
c) his comments were taken as his personal view of women
d) his comments were construed to explain the low rate of women getting tenure at Harvard
e) only one or two of the conference participants actually took offence, and that seems to be because they believed c) and d) above. Everyone else (who was actually there, unlike the preponderance of pundits who commented on the situation) thought it was totally unfair to criticize what Summers said.
Needless to say, it was an unfair witch-hunt blown completely out of proportion by the fact that he happens to be the president of Harvard. I agree that he was stupid to think that he could somehow decouple his academic persona from his administrative persona, and that he probably mangled what he meant to say and came off sounding terrible. But I also think that the conference organizers should be ashamed of themselves - if you get Summers to come by telling him it's off the record, it should be just that - off the record. It's shameful that the Globe was able to report his comments at all and had any sort of access to the proceedings.
My story: when I was in 7th grade (Year 8 at good ol' Chinese International School in Hong Kong), my maths teacher sat my parents down and told them very gently that I was a very nice girl but that I was terrible at maths - not only could I not mentally arithmetize my way out of a paper bag, but I didn't know any of my times tables. The other children had been sent to Kumon every day after school and had memorized all these things, and I was far behind the class. Perhaps she should concentrate on English and reading? my teacher said kindly. My parents went home furious, but found to their chagrin that the teacher was right - I was indeed terrible at arithmetic and the times tables. However, the next semester I was taken in hand by a wonderful Hungarian man named Mr. Lazlo Varro (who I have never been able to trace to thank for his faith that anyone can learn to do maths) and taught calculus. I beat the pants off all the kids for whom maths was memorization and not actual understanding, and now I'm at Harvard having finished my statistics degree (yay electives this semester).
Conclusions: many girls can do maths very well. Some can't. Some boys do maths extremely well. Many don't. However, schools shunt girls out of maths to concentrate on the small percentage that do extremely well, because that's how teachers are evaluated, and also because girls are routinely pushed out of maths at all levels. I think girls need more encouragement in general to do stuff, although I don't know if this is a nature or a nurture thing. A study I read last year showed a similar imbalance in women seeking office - once women actually run for office, there's no inherent bias against them, but many fewer women even bother to run initially because they are far less likely than men to get support and endorsement from political parties. This is a very interesting study, and well worth a read (Lawless, Fox. "Why don't women run for office?" Taubman Center for Public Policy, Brown University, 01/04, and shockingly, it's online in pdf format.)