You guys are vastly oversimplifying this whole thing. The only reason JSTOR gets access to these documents in the first place is because of licensing agreements with the publishers.

You can try and a start a company that gives it all away for free, but you're not going to be given access to any documents or you'll be sued for releasing the ones you don't have rights to.

This is not as simple as "Just throw that stuff up on a Torrent and we're good to go".