The question is: why would female participation in CS drop by half from 1985 to today? Why would innate preferences change so dramatically in that time? Certainly, the usual characteristics trotted out for why CS isn't attractive to women were even more true in 1985.
I don't think it's entirely fair to say that female participation has dropped since 1985. The raw number of women has not, in fact, dropped. In fact it has risen, it just hasn't risen as fast as male participation.
