I wonder if chrome was counted twice? The fact that some are still running 11 and other were running 12 makes me think that maybe users were in the middle of that switch? I'd put their numbers around 12% then.

Whatever it is, I think it good. My app is biased to chrome because that's what I use. I keep seeing more and more people on chrome at my browser research lab (snooping on people's laptop at starbucks ;p)