The repeated use of "stole" in the indictment is interesting, even beyond the usual metaphorical usage to discuss copyright infringement.
In this case, the indictment alleges that the documents were stolen from JSTOR, which does not even own them! In the vast majority of cases JSTOR scanned documents whose copyright is owned by someone else, and acquired or was donated a non-exclusive license to distribute copies via its service. In many cases the documents are even public domain. The indictment continues the theft metaphor by discussing the effort and expense JSTOR incurred in scanning the documents, and the alleged attempt to render this less valuable by redistributing "its" documents, analogizing this to the loss someone suffers in a theft.
But effort expended to build a private repository consisting of copies of things you don't own doesn't give you ownership of the result, any more than Google Books doing the same has given them ownership of the documents that they've scanned. If you scraped Google and "stole" their scans, you would be violating Google's Terms of Service, and Google might indeed feel subjectively like you've taken something of value (their exclusive access to this repository of scans), but I think it would be a stretch to say that you've "stolen" "their" documents.