And what has the actual result been over time? Are we looking for confirmation anecdotes in the wrong places? It seems to me that many of those early stats turned out to be true. For example, Android devices dominates the global smartphone market at over 80%. That's either a lot of shipped but not sold devices sitting on shelves filling up every inch of floor space in retail stores around the world, or this kind of stat tracking has some validity.
It would be easy enough to test the location theory: just slice your web traffic by country or find someone with a big non-US site
Unless something has changed, you still end up with a data plan which is expensive if you don't plan on using it at all. I suspect something else is going on with the reported stats, but I'm not sure what it is.
I'd really like to see some more in depth research on this phenomenon whatever the cause is.
StatCounter relies upon subject sites adding a bit of code that reports back to StatCounter, giving them their metrics. Curious, I inspected every single site I visit in a typical day, and zero of them have the StatCounter code.
I am invisible to StatCounter. I don't roll with Chromebook, so not really relevant in this case, but it's interesting how frequently StatCounter is used to make some proclamation or other with no one ever questioning what their coverage is.