Of course, their definition of what constitutes a "junk political propaganda site" is entirely subjective, so this is a completely meaningless study. I consider CNN to be junk political propaganda, how do the statistics vary when you include THAT? If they really wanted to make a meaningful study, they would publish statistics for several definitions of "junk lists", collected from varying perspectives. I suspect that would just be another way to show the stark polarized digital corners we've all painted ourselves into. As is, this is, ironically, more political propaganda....

Perhaps you ought to take a look at this patent submitted recently: http://www.wipo.int/pctdb/en/wo.jsp?WO=2008039176