Look for links in the phrase "such scans are often considered uncopyrightable" in the original article for more on this topic.
We've discussed this on HN before. Reproductions of public domain works are not protected by copyright.
Edit: Of course a significant portion of JSTOR's archive does consist of copyrighted works.
So I think you are correct.
>You are free to copy and use individual images but copying and distributing the complete collection may infringe what is known as the “collective works” copyright. Collections of public domain material will be protected if the person who created it has used creativity in the choices and organization of the public domain material. This usually involves some unique selection process, for example, a poetry scholar compiling a book -- The Greatest Poems of e.e. cummings.
I don't think JSTOR is using what could be described as a unique selection process.
Let's say I come into possession of a trove of public domain documents from the 1800s, and I take the time to scan every one of them painstakingly and from those scans to assemble an academically useful database, for which I charge $100/person/year for access. Your contention is that because the underlying documents are public domain, you are entitled to unfettered access to my work product, despite the fact that I took steps to gate access to my work product.
The law isn't going to see it that way, I don't think. But maybe you're right. I'm not a lawyer.
Like everyone else on HN, I wish all these documents were freely available too.