Perhaps. This woman argues that the differences are self-exaggerating, that fields which fewer women are interested in pursuing tend to be male-dominated, which makes them even less attractive to women, which makes them more male-dominated, in a cycle which leads ultimately to a situation where only the women most devoted to the field stay in it.
But that's taking the argument one step too far already. If the question was why the situation at Google is 80-20 one needs only to look at the graduation statistics from US Comp Sci PhD programmes (if Google hasn't changed their hiring practices recently), where the figures are indeed in that range (and that's counting mostly foreign women, without them I seem to remember that the figure would be closer to 90-10).
Now you're already addressing the question of why women have opted out much earlier in the chain, and while that is interesting, it's not really something that Google can fix with their hiring practices, they can only hire from the candidate pool that is there after all. It takes something else. In another part of society (of which admittedly Google is a part, so they can do something, of course).
Now, I know from first hand experience in the academic teaching field how unpopular it is to (as I've had to do) point out that our targets and goals of increasing comp. sci. female undergraduate admissions were completely unrealistic as we would have to attract (in that case) all qualified girls from high school, not a single one would be left for medicin, law, etc. which we know already attracts a majority of the qualified female students. But like the aspie idiot I am I feel it still needs to be pointed out. (And I have tenure, so I'm harder to fire... :-))
Our answers in both acceptance and hiring to the "WHY DON'T YOU X MORE WOMEN" (where X is hire/accept) is and continues to be, "because they aren't there and they don't apply". We can't fix that at the end of the pipeline. (And I've been exposed to that in both industry and academia for more than twenty years, no come to think of it, it's closer to thirty...)
And being in Scandinavia I'm not sure I buy the "there are too many men there" argument. Thirty years ago that was very much true of medicin, veterinary medicine, and law to mention just a few highly sought after careers, difficult to get into and more importantly almost 100% male. And today Swedish universities have been e.g. fined for instituting "affirmative action" programmes for boys so that the veterinary programme (or was it law?) wouldn't be completely female. (But that's against the law, so no boys in that field...)
For example, in 1992 (Sweden), medical doctors 55-62 were 93% male 7% female. In 2010 in the youngest cohort it's the other way around, with 39% men and 69% women. If the "(old) men scare away women" hypothesis would be true, then this change of affairs is a very clear (data) point against. At the very least it didn't work on doctors.
Or lawyers, 57% of all judges in Sweden are women now. 55% of all judges in criminal matters are women, and that's set to change even more, as their dominance in the younger cohorts are ever more marked. If not even the grumpy old judges managed to scare the dainty young women away, well, that's another pretty hard blow against that hypothesis. (That doctors are wishy washy and can't put their collective foot down is after all somewhat believable, but scary and scarred judges, well they were kind of our last hope! :-))
But of course in comp. sci. the figures are pretty much identical to what they were in the eighties. There are a few more now, but we haven't nearly have the sea change that we've had in medicin (both veterinary and human) and law.