Yes, they are doing that themselves. But hopefully not for much longer if the internet has anything to say about that. I applaud all of the scientists who are working hard to make their research more open.
Imagine a world where Swartz managed to destroy JSTOR. A professor asks his grad students to read an article. They can either dig it out of the archival stacks and read it there or wait for interlibrary loan or...use a torrented copy or illegal DVD they bought on the street at their last academic conference. There is no way to track which one they did. No one will ever know. And no one will care: all the prof cares about is whether his students read the damn article; he doesn't care how they got it. He's completely focused on getting his research done, writing grant proposals, etc.
He wanted to provide access to the articles to people that do not have 50K/year to pay to JSTOR. That would by no way destroy JSTOR
JSTOR prorates its fees so that poor institutions pay much less than rich institutions. And all of Africa gets JSTOR for free. Now, even individuals can get capped free articles from JSTOR.
If the JSTOR archive was released to the world, why would journal owners continue licensing their content to JSTOR?
Finally, do you understand that most of what JSTOR charges pays for either licensing the journal content (JSTOR doesn't own copyright for the journals it includes) or for scanning/digitizing new content? The costs that JSTOR charges pay for a hell of a lot more than a few web sites pushing PDFs to clients.
I'm not sure how this question makes sense. If the content that is accessible via JSTOR would be also accessible otherwise, which journal owners would withhold permission from JSTOR to publish their content? That's like saying if the book is available in the library, nobody would ever buy it. Obviously, it is not so.
Also, surprisingly, in September 2011 JSTOR made public domain content available to everybody and shortly after that instituted free individual access programs - which they didn't do before, and which by your words should be impossible due to high costs that can not be supported in any other way but by charging for any access. Somehow it turned out to be possible still, and JSTOR did not collapse.
I wonder if Swartz, who was very concerned about information access in the global south (according to his manifesto), knew about that.
And spoilers: it's not. In order to make it would you would have to recreate everything JSTOR currently does, which would result in a JSTOR not called JSTOR. One with legal liability at that.