Any time I've seen an article from The Guardian (or Washington Post) in the past 7 months, it's been Snowden related. If the poll was conducted in some other newspaper, maybe the distribution would be different, and there would be more than a meager 2,200 votes.