Beyond what has already been mentioned by people commenting on this "study", the biggest problem I can see here is that he doesn't seem to be accounting for the size of the blogs at all. DailyKos, for example, has a huge member base, and every entry gets hundreds of comments from readers. It seems to me that a much more scientific way of doing it would be to calculate a "profanity per capita" measure, because what his study may show more than anything else is the fact that there are a lot more people frequenting left-wing blogs than right-wing blogs.
That could be a good exercise. Since I don't visit most of the sites he has listed, I have no idea how many people there are using those sites. You should recommend the author to try it out and see how it changes the figures.