I just wish there were some way for it to deal with paywalls -- since most of the pdfs I need are behind JSTOR or Muse. I realize this would be a technical and/or legal hassle (things from behind a paywall would have to be private by default, I guess) so I'm not holding my breath. But for use in the humanities and social sciences, it would make it a lot more useful.

I keep meaning to write some script that downloads jstor PDFS that I want to pin, uploads them to my server, and bookmarks them as private there. But trying to figure out whether I'd need to add a robots.txt and the like has kept me from making it happen.