The way I view copyright law (IANAL), this is the correct approach. Each upload of an item to Megaupload is an independent assertion by the uploader that they have a license to put the work in question on the site. _Content_ itself is not prima facie infringing; instead, it is the unlicensed _use_ of the content that is infringement. This was essentially Youtube's argument when Viacom uploaded "roughed up" content to their site [1]. In that case, the roughed up copy as originally uploaded would be a licensed copy, but any copies that other users uploaded, even with the same MD5 hash, would be unlicensed.
On the other hand, child pornography can never be "non-infringing." It is always illegal to knowingly store obscene content, regardless of who stores it. Thus, one link to a child-pornographic file is as culpable as the next, justifying mass removal. This is essentially the argument made by Katherine Oyama of Google at one of the SOPA hearings [2].
On the other hand, child pornography can never be "non-infringing." It is always illegal to knowingly store obscene content, regardless of who stores it. Thus, one link to a child-pornographic file is as culpable as the next, justifying mass removal. This is essentially the argument made by Katherine Oyama of Google at one of the SOPA hearings [2].
[1] http://news.bbc.co.uk/2/hi/8575666.stm
[2] http://danwin.com/2012/01/the-sopa-debate-and-how-its-affect... (Ctrl-f "REP. MARINO")