So according to your argument, the risk is not that we will be unable to access such articles without JSTOR, but that we will be unable to search for articles?

So according to your argument, the risk is not that we will be unable to access such articles without JSTOR, but that we will be unable to search for articles? That's basically saying that a document search system is too costly to create or maintain for a university library to deal with, and so JSTOR is necessary for queries. Which sounds like an unfounded assumption to me, considering how widely deployed CiteSeerX is. Even if your argument were true, what exactly was Aaron draining? If JSTOR is providing this valuable search service, wouldn't that alone fund their operation? If having millions of documents is useless because we have no way to search those documents, what difference does it make if Aaron really had been sharing the documents with others? Would it not have been a good thing if Aaron had amassed these articles and made them available using a better search service -- wouldn't we have benefited (and isn't that the whole point of copyright anyway?)? Had Aaron been accused of hacking into JSTOR to download the source code of their search system, you might have a point. Instead, he was accused of copyright infringement and of violating a network use agreement (because he was trying to evade the ban of his laptop's IP address); at no point was JSTOR's valuable search service even an issue in this case.