Yep, but that's based on their spam filtering service. I have no idea how that corresponds to the overall mail "population", but I'd expect to see a higher hit rate on such samples as by definition people who pay for spam filtering already have a spam problem.
I'm not saying there's not a lot of spam out there, just a bit skeptical about the accuracy or usefulness of those sorts of statistics (apart, of course, for marketing anti-spam services :)
Nik> Agreed. I posted it as many people struggle to find interesting security stat quotes.