To be honest, I mostly wanted to comment to share my anecdote and make a Prince reference. I'm not really trying to make an argument one way or another for browser popularity.

BTW, I personally think the browsing-time metric is better than unique-users, because applying the latter metric to music would have had Rebecca Black as the most popular artist of 2011. Which conclusion, though true under some definitions, might not be quite as useful as other possible conclusions.