I wouldn't want to be the one arguing that case.

Sure, some artists used Megaupload to share their own works, but these uploads were far outnumbered by their illegitimate counterparts.

Has this been demonstrated yet? (I currently slightly favor your belief but I could see it being the other case.) I know that Megaupload provided tools for the labels et al. to remove infringing content in a very easy way (just like Youtube offers), and there was a lot of outcry from people over their personal files being lost because they were treating it like dropbox.

When called into question, users will argue these platforms/protocols can be used for legitimate purposes, and point to examples of people using them as such. In this case it was Megaupload attempting to do this to defend itself.

It doesn't, however, change the reality of the situation.

The more interesting case to me is whether one paints a service on distinct counts of items or on total traffic per item. PirateBay makes it easy to see what the hottest torrents are (http://thepiratebay.se/top/all); I'd be surprised if any are authorized. If we assume a power law then most of torrent traffic certainly infringes. For Megaupload, perhaps it was the same. But I'm more inclined to think Megaupload's distinct counts fall closely in line with Youtube's--it's trivial for copyright owners to take their content down with either a DMCA request or the tools the sites provide, meanwhile tons of user-generated content is appearing that actually belongs to the user. MediaFire seems like the place to go these days for sharing your songs, even professional artists use it sometimes. I don't think it's that much of a stretch to consider that for Youtube and perhaps Megaupload, there is more non-infringing content than infringing in absolute terms even if the infringing content takes up the supermajority of traffic.

It's interesting you point to The Pirate Bay's list of most popular torrents though. The Mega indictment cites an email from one of the alleged conspirators to another employee instructing said employee to alter the Megaupload 'Top 100' list. Elsewhere it is alleged that this list did not accurately reflect the most popular downloads on the site, and that this was done deliberately to make the site appear more legitimate than it was.

Not sure why I'm being downvoted, but the simple fact is that there was enough evidence to have an indictment written up, signed off on and then authorities in another country agree it had enough merit to be acted upon.