Part of the reason that I don’t think your paragraph 1 assumption is flawed is that there is evidence that men and women were equally interested in programming before the 80’s, but that a combination of cultural and societal factors came into play in that decade that branded computers as a “boy thing” and resulted in women being pressured out of the pipeline from a young age.
But even if you can find some sort of counterargument for that, my overriding reason why I stick with this assumption is as follows: decade after decade, people insisted that women weren’t into $career. And one by one, women have found their way into those careers (factory work in WWII, the executive suite more recently, plenty of other examples in between). In each case, the people saying women weren’t into $career had an argument that seemed to make sense, often relying on some combination of status quo statistics and hand-wavy conclusions drawn from psychology. And they were wrong. So when I see those same arguments brought out again for this career (often with some exhortation that “this time is different”) I generally dismiss them, as the epistemology they’re based on has a bad track record.
With all that being said, when it comes to actions, I’ll be more likely to encourage women to get into programming if I get a sense they might be interested in it. However, when it comes to interviewing candidates, I will not grade people on a curve based on their identity, as it is my belief that men and women can in fact be equally good at programming and should be held to the same standard.