Thanks for your insights.
I do not think anybody would assume that a maths tool is representative of uptake by a whole country. But it does allow to confirm the realism of increased W10 and CB use. Basically, if several factors and stats point into the same direction: the likelihood increases that the observations are correct.
I do think it is a valid check of generally increased use of Chromebooks within the user-population and that they are mainly NZ-based.
The switch from W7/8/8.1 to W10 is also not unexpected. It supports the idea that the stats are accurately collected and counted.
I would be interested in Linux, ChromeOS uptake in general. With that I am automatically interested in the trends of the main competitors: MacOs and Win. Uptake of Linux is so small that an accuracy of 1% can leave many stats quite uncertain with regards to linux. In countries where CBs are at levels that MacOS was a decade ago (i.e. 4-5%) the picture about ChromeOS is becoming clearer.
In countries like Sweden with Win below 70% uptake and Mac at 25% and chromeoS at 5% & a combined with a good connectivity infrastructure, the call for cloud-software is understandably bigger than in many other countries. It becomes less of a default-choice to assume applications must be Win because everybody has it.