But what's the mechanism for how it can ever actually do that? Suppose there was discrimination that meant some women who "should" rightfully have done CS degrees instead did something else (and I don't think anyone's ever actually shown this without making an arbitrary assumption that any difference in the number of applicants must be due to discrimination, but let's put that aside for the moment). So now you have a number of women with less CS experience than they rightfully "should" have. If you lower the bar for women to give conference talks, or get promoted in the workplace, to compensate for this lesser experience, you're not actually filling that experience deficit, you just get a number of women who've been promoted above their experience level. That doesn't fix past discrimination, it makes it worse.