The curious part is that they allow web scraping arbitrary pages on demand. So if a publisher could put in a lot of arbitrary requests to archive their own pages and see them all coming from a single account or small subset of accounts.
I hope they haven't been stealing cookies from actual users through a botnet or something.