I've seen this mentioned elsewhere, but I can't see anything on the Statcounter site that confirms it.

http://gs.statcounter.com/#mobile_browser-ww-monthly-201006-... breaks browser usage for iPod and iPhones out explicitly, and the sum of those both is less than the total iOS usage in http://gs.statcounter.com/#mobile_os-ww-monthly-201006-20110...

I could be wrong, but if I am then it would be a weird thing to do - it would mean that Statcounter doesn't list iPad usage anywhere (eg, the other browser-by-version graphs exclude mobile Safari)

For the last three days: iOS = 19.98%, iPhone = 15.14% , iPod Touch = 4.76% , iPhone + iPod Touch = 19.90%.

The difference is only 0.08% of mobile usage, so either they don't count iPad usage as part of Mobile at all and that's just a rounding error, or they count it as drastically lower than the 1% (of all browsing, not just mobile) that other sources are quoting for iPad browser share.

Actually, you can tell they do count iPad as part of the normal, non-mobile, browser graph. It's small enough that it gets bundled into "Other" on the browser versions graph, but if you download the CSV data it's called out separately, currently at 0.73% of all browsing and the total they give for all versions of Safari combined wouldn't add up unless they counted the iPad versions in with the desktop versions.