I'm still planning a post about author gender issues in general, but for now I just want to mention one particular thing.
Cheryl and Nicola Griffith suggest (due to statistics compiled by John-Henri) that (except for a couple of recent years) the percentage of female winners of Hugos for fiction has been steadily increasing for decades, except for a recent small dip.
Unfortunately, I think the statistics may be a little misleading.
First: of the 12 fiction Hugos won by women from 1991 through 2000, 6 of them were by Connie Willis, and 3 by Lois McMaster Bujold. I'm certainly not saying Willis's and Bujold's wins don't count; of course they count. But it's worth noting that at any given time, there have often been one or two female sf writers who get a lot more attention than any other female sf writers. (For example, there are a fair number of old Year's Best anthologies featuring ten stories by men and one by Le Guin, with no other female authors.)
Thus, the total number of women who've won is smaller than the percentages make it look. (On the other hand, male authors often win Hugos multiple times over a span of years as well; Haldeman and Resnick each won 3 during 1991-2000, for example. I haven't looked at the percentage of female vs male winning authors per decade.)
Second: I also think it's worth looking at the number of nominees on the ballot rather than just at the number of winners. (Though number of winners is certainly worth looking at too.)
In particular, as I noted in a comment on my abovelinked journal entry, in 1992 and 1993 there were ten fiction works
per by women nominated for a Hugo each year. [Typo "per" corrected later in the day; sorry for any confusion.] For most of the rest of the '90s, the number each year was 6 to 7. Over the past ten years, it's been more like 3 to 4, with possibly-fluke local minima (of 1) in 1998 and 2007.
Sue Linville's graphs of stories by women in Asimov's and F&SF also peaked in the late '80s and early '90s.
So my conclusion tends to be that female authors got a lot of (well-deserved) attention in the late '80s and early '90s, but that that attention has (unfortunately) declined somewhat over the past 15 years.
Looked at from that angle, John-Henri's statistics seem to me to suggest that, except for an unusual peak in the '90s, the percentage of Hugos for fiction given to works by female authors has been fairly steady at roughly 25% since the 1970s.
The Nebula percentages do look more clear-cut, but I haven't looked at those numbers in detail.
. . . I should note that it's always dangerous to try to draw statistical conclusions from too-small sample sizes. Which is presumably why John-Henri aggregated by decade--but in the Hugo case, I think that aggregation obscures some important details.