Hint to the uninformed: don't look up "bitmap" and wonder why anyone would index them. Look up "bitmap index" instead. This will save you many clicks in the github project as well as in your favorite search engine. It's about databases, not pixels.
The key addition with Pilosa is that it's distributed and can scale horizontally :)
Doesn't this also make it several orders of magnitude slower? Every time I find myself using Bitmaps it's when speed is extremely important. What are some use cases for this not covered by in-process Bitmaps, Bloom filters, and HyperLogLog?
I'm just not aware of many use cases for bitmap indexes willing to trade that much speed for database-like access. I think this project would be more usable factored into a library than something external
At my company we also strongly use Bitmaps on the analytics database engine we developed (S1Search).
I think you're right though, that there are use cases which would benefit from a library exposing this functionality - you need to have quite a lot of data before compressed bitmaps representing the relationships in that data start overflowing memory on a single machine.