There was no sharp drop-off in the 80s. What happened is that interest from men soared while the proportion of women choosing to study it remained relatively constant in comparison, only dropping by a little in terms of the actual percentage of the female student body choosing it over other subjects. Looking only at the ratio is completely misleading.
There was only a short exception around the PC boom and the dotcom boom, when the proportion of women choosing to study it actually increased, and then fell down again once the boom was gone. This is similar to what the Scandinavian countries found in trying to encourage women to go into engineering: when they stop the campaigns, women's participation drops back to baseline.
But rather than look at the complete picture, only the gender ratio graph is ever shown, because that's the one that confirms the preconception that women are being victimized. This "drop", by the way, only exists at the bachelor level, and vanishes at the masters and PhD level, which suggests there is nothing driving committed and talented women out of the field at all.
But even if you didn't think all of the above is valid, it is preposterous to posit that the biological differences argument implies women must've somehow changed in the mid 80s. There is a much simpler explanation: computer science itself changed, and how it relates to industry. With the birth of PCs and especially SaaS, the amount of hard CS skills required to make computers do useful things dropped pretty much to zero. Only people with a particular theorical and abstract interest need to go into it. At the same time, software development pioneered many of the async and remote working styles that are common place in open source, and which allow for individualistic collaboration rather than the always-on social setting that women prefer. You see this in engineering too. Women gravitate towards engineering subjects that feature more team work, field work and lab work, and which involve working with people. The increased appeal of the field for men could be due to the fact that a segment of people previously tortured by social styles could now excel in a more solitary setting.
None of this is a taboo when it is used to imply that women's communication and empathy skills are innately superior, a common refrain in diversity circles. But try to make the same argument about men and their abstraction skills, at a population level, and suddenly you're blaspheming against the church of intersectionality.
The double standards and shoddy science are obvious, and the NLRB should be ashamed for rendering a judgment that declares reality to be sexist because some people's feelings were hurt. The only person who got actually victimized was Damore himself, for the crime of taking diversity advocates at their word, and actually offering them the feedback they claim to want, but clearly are incapable of processing.