I think Debbie hits the nail on the head with her comments and her questions. As someone who works on this topic, here are my attempts to answer:

1. I think the experience of women in computing is slightly anomalous in that women were able to remain in the field in relatively high numbers in the US and Britain once the field had already acquired a certain level of cultural prestige (and when the jobs were quite lucrative). On the other hand, the pushing out of women from a field as it professionalizes is a very common story. So computing is not highly anomalous in the history of gender and labor. I think it just seems so because there was a "delay" so to speak in women being pushed out, because of the work's feminized, supposedly deskilled origins. I think it also seems anomalous because we (Americans) have been conditioned by the media telling us for ~20-30 years that computing is aligned with (white, middle class) men's interests, capabilities, and accomplishments. So it still seems strange and new to many people when, say, Walter Isaacson "discovers" women in the history of computing.

2. I think that yes, women in computing are a bellwether of sorts. Their gradual removal was a harbinger of the fact that computing was becoming less about computing and more about management. (I resist calling it "flight" or saying they "left" because that is an inaccurate way of describing the changes wrought by structural discrimination.) Women were fine when the work was seen as technical; not so much when the work became aligned with figuring out how to wield (managerial) power most effectively. I also think women will be a bell wether in the reverse: the more women we get into programming, the more likely it will be a sign of the field once again becoming seen as deskilled or less skilled. The abuse of H1-B workers in the US is a harbinger of that trend.

3. I am sure the experience of women in computing is quite different from one country to another, but I'll just talk about the differences between Britain and the US, since that's my bailiwick. In Britain, the rationale for actively chasing women out of computing in the 1960s (except in times of labor crisis) is very clear, and it comes directly from the top--i.e. the government. It's a clear example of the nation's largest employer deciding that computer work of any sort (operating, programming, systems analysis) was too important to be left to supposedly unreliable, lower-level workers who did not have management potential, and who were more aligned with labor than management. It was an issue of both gender and class (with class itself being a gendered category). This seems different than the historical literature on the US context, where there is more focus on how discourses of gender affected people and (it seems to me) somewhat less focus on very discrete, large-scale cases of structural discrimination.

I think your insights and questions also resonate with the concern I voiced earlier, Debbie. In paying so much attention to the seeming tragedy of women in Anglo-American computing over the past few decades, and trying to produce historical answers that can be translated into solutions for women's underrepresentation today, are we in fact strengthening the systems of privilege that created this problem in the first place? I often worry about being a participant in that, given my work on this topic.

Best,

Marie