Except that scraping itself would be prohibited by copyright law. This is why there should be some central repository, so that there can be a legal copy that is available to the public after the copyright expires.