Personally, I still think this is the most parsimonious solution to the question. Its only flaw is that it doesn't give anybody any reason to self-flagellate, or at least not much of one.
People keep trying to have this discussion in the context of the 1950s, but we're in 2014, where women are, among other things, the majority of college graduates (and increasing). It is a obvious fact that in the past 20-30 years, every discipline has made a huge push to attract more women, with great success. The real problem today is that computer science/programming needs to attract women in a world of incredibly stiff competition.
Why is enrollment declining? Is it that unreasonable to suspect that the answer is "because computer science is being outcompeted by the dozens upon dozens of other disciplines making women better offers"?
I deliberately phrase this as "computer science" and not "computer scientists" because I'm unconvinced the problem is in the people. "But sometimes, there are misogynists in the computer world"... how is that unique? Every field had that problem. Today, computer scientists have been damn near actively begging for women, to the point that the solicitousness itself is becoming offensive, as the author points out. There gets to be a point where advocacy turns into trying to tell women what to do.
Unless both genders are precisely equally drawn to every single discipline, an idea which I think is itself rather offensively insensitive ("feminism" shouldn't be defined as "women are just funny-looking men" but my goodness is that a popular version if you really look at what people believe), there's going to be unbalanced disciplines.
This goes against some of the doctrines, and so I think people tend to put this argument when I make in the wrong bucket, but personally I think the question of whether we're trying to tell women what to like and how to like it is itself a feminist question. "So we should just stop trying because it's hopeless, huh?" No, I don't think that. Roll the carpet out, with all due honesty and openness, sure... but if nobody chooses to walk on it, that may very well simply be their decision (in aggregate). We have no right to run up to them and tell them they are wrong.
Might I also add that by sp332's own statistic, that the ratio has been falling since 2001, the idea that the field has been getting more misogynistic in the past 12 years is just patently absurd, and that is itself evidence that what's happening here isn't about the people.