That point is exactly the first thing I thought of when I read the article. How much of Tumblr's traffic is for "photo blogs" of naked girls? Based on what I've observed, probably a good amount.
Not that that traffic is somehow invalid, but I think it makes the comparison a bit more apples-to-oranges.