Earlier quoted context omitted.
You just gave me a sequence of data points that confirms my argument, claimed instead that it contradicts it, and then linked to a rambling SlateStarCodex post.
Not really, you claimed double digit differences, which is incorrect as shown by the data. You claimed widespread prejudice, which has no supporting evidence. And you may consider the article I linked to be rambling, but it's comprehensive.
The "people / things" role postulate is interesting, it may explain some of the differences. But I don't think it holds up in all cases (50% of chemist bachelor degrees are female (https://www.acs.org/content/dam/acsorg/membership/acs/welcom...) and chemistry really isn't a "people" oriented discipline IMHO).
In general culture, I do think some things get grouped into one sex or another based on pure marketing and image. The marketing style or image itself might play on certain characteristics of the sexes that are biological (for instance, men have more testosterone of course, so men will respond better to marketing and imagery that plays on testosterone oriented characteristics). But this might say nothing about the product itself.
For instance, I see nothing biological at all why in most Western societies, beer tends to be seen as a "masculine" drink and wine a "feminine" one. Rather, to me it seems to be pure marketing positioning at this time.
With CS, there may be some biological explanation which will produce a natural bias in the ratio. But there may also be a marketing / image / "role" component of CS that does depress the ratio as well. IMHO, the marketing / image part of this is always worth challenging.
And there is an ingrained stereotype with computer programmers: the popular image of someone into computers in Western media is, pretty much almost always, a socially awkward, non-athletic, nerdy male. (This stereotype honestly is actually honestly unfair to male programmers that aren't socially awkward or are athletic or aren't terribly nerdy.)
It would be interesting to examine the popular stereotypes and generalizations of programmers in other countries and see if sex ratios differ based on what the positioning is.