> Why is that a flaw? Why should the numbers include those devices?
My guess is that he's trying to compare OS vs OS, as opposed to OS vs device. Do these #s include traffic from Android tablets or MIDs? (Are there even any available for purchase at the moment?)
It's a (currently unique) advantage to the iPhone that it has a very popular sister device in the iPod Touch that is interchangeable in most respects e.g. designing a website, codec support, App market size.
Ignoring it would be the equivalent of comparing iPhone sales against a single Android phone (which I've also seen done repeatedly) which ignores the Android advantage of being available on multiple phones from multiple manufacturers.
Almost any actual decision you'd want to make about iPhone or Android as competing platforms will need to take into account the existance of iPod Touches or Android phones by a variety of manufacturers.
The numbers in each case may be correct for the limited arena they cover, but they're still misleading for any realistic purpose.
Why is that a flaw? Why should the numbers include those devices?