Just the other day, I delved a little more deeply into the Internet Archive’s collection of “ebooks” … and discovered they have managed (?) to upload close to one million texts with “google” as the scanner, and have an option to download for Kindle. Two things: these are apparently uploads of the source PDF image scans, and have been re-OCR’d; and the mobi conversion seems to break fairly frequently. The OCR is definitely of inferior quality to the Google scans, which are of course what are in the Google epub downloads, and lack (for example) structure such as table of contents. Additionally, with RetroRead, you can edit extraneous front matter imagery; full-text search the library using Google search; and of course, easily download direct to your Kindle.
… Let a thousand flowers bloom.