[anologwintermut]

Summers's point was about the distribution at the extremes, not the average. He was addressing the lack of professors in STEM at places like Harvard which manifestly select for the far end of the bell curve. Right, wrong, or otherwise, that has almost nothing to do with the average case and even he contended that the best statistical evidence showed women and men were roughly equal.

So, unless one seriously thinks that the entire field of programming/IT/computer science as a whole requires that level of talent, Summer's point doesn't apply and there are certainly other reasons for the gender gap in computer science.

[Trumpet6]

A survey (published in Science[1]) of ~1800 academics across 30 disciplines in the US saw a non-significant positive correlation between estimates of how selective a department is with graduate students and the female representation in that disciplines.

In other words a field that was more selective (as measured by the proportion of applicants being admitted) might have more women in it.

This evidence is is not compatible with any reasonable prediction from the "different high tail" theory.

[1] http://www.sciencemag.org/content/347/6219/262.abstract