That would only be true if the most popular files are pirated, if the most popular files were legal that'd skew it the other way. So for this to even be an argument the figure has to be >50%, but possibly higher than it should have been.
I'm not sure how you come to that conclusion. For example, suppose there are a hundred million files, a million are pirated and have on average a million downloads each, whereas 99 million are legitimate and have on average a hundred thousand downloads each. 1M * 1M = 1B. 99M * 100K = 9.9B. So in this example, only 1% of the files and ~9% of the downloads are pirated, but using this study's methodology the false conclusion would be that 100% of them are pirated because the pirated ones comprise the entire top 1000.
And this is in the nature of what actually happens, because Hollywood films are popular -- you can pretty well expect that The Dark Knight is going to get a hundred million downloads, but that tells you nothing about the potentially quite large number of legal downloads available which are still getting 100K or even 10M downloads but not enough to make it on the list of "most popular" files.
You don't show that a better study would come to a different result, you just claim the proof is so weak it's worthless.
That's what debunking is. You show that the conclusion doesn't follow from the premise. If you want a different, actually valid conclusion then you have to run the study again with better assumptions.
What you're doing is assuming the conclusion without proving it and then, when someone points out that you haven't proven it, retorting that they haven't proven the contrary. You're the one asserting that 99% of BitTorrent is piracy, so where is your proof? We've already established that the nonsense study is not it.