[This was my first stab at answering why there is so little femslash, and looking at stats on the topic; more possible reasons follow in the next chapter.]
I suspect one of the main reasons is the lack of source material with at least two strong, well-developed female characters. As Linda Holmes at the Monkey See blog recently observed, At The Movies, The Women Are Gone (which she backs up with stats at the link — go read it; it’s fabulous):
I want to stress this again: In many, many parts of the country right now, if you want to go to see a movie in the theater and see a current movie about a woman — any story about any woman that isn’t a documentary or a cartoon — you can’t. You cannot. There are not any. You cannot take yourself to one, take your friend to one, take your daughter to one.
There are not any.
This pattern — not limited to movies — severely limits the number of interesting potential female pairings for fandom to play with.
If I’m right that this is a major factor, then we should see more femslash in fandoms of media that have more well-developed female characters. I decided to do a brief analysis of a few stories focused on women and their relationships to each other, and of a few popular ensemble cast shows with at least two strong female characters. I also tossed in AO3 as a whole and Sherlock (a show low on female characters [especially at the time of the original analysis]) for comparison:
You can see the numbers in the spreadsheet.
Keep in mind that the number of works in some of these fandoms is very small — there are fewer than 100 works apiece for Bridesmaids, Anne of Green Gables, and My So-Called Life — so the exact percentages are not terribly reliable. Still, we can see that having a higher number of well-developed female characters and female interactions definitely correlates with an increase in femslash.(At least among the media I’ve chosen here, which were generated somewhat arbitrarily via my own brainstorming.) [However, the fandoms highlighted in Chapter 1 as having a high proportion of F/F fic seem to back up my hypothesis.]
I don’t think that’s the only factor at play (I’m also fascinated by all the queer women writers out there, myself included, who mostly focus on M/M pairings), but I suspect if we had an even playing field to start with, it would drastically change the ratio of femslash to other relationship categories.
Also, a quick historical note: according to another person who responded to my series (whose note I can’t find right now — so sorry to not give credit!), AO3 was founded by people mostly producing M/M slash, and their friends and readers were the first to start populating the archive. There’s a decent chance that this initial population bias has remained as the archive has grown — resources that get used mostly by a certain community will become well-known within that community, and attract more members of that type (in this case, M/M writers). So AO3 may be more biased than fandom overall. I don’t know of any other archives that let you search by relationship category in the same way (FF.net doesn’t seem to), so I can’t immediately test this theory.