To be fair, isn't measuring rate of growth sort of misleading when it comes comparing large bases to smaller bases?
In other words, a county with 1,000 people would grow by 10% if only 100 additional people moved in, while a county with 100,000 would only grow by .1% is an addtional 100 people migrated to the area.
I'm not saying that the disparity of the install bases of Windows and Linux is as great as my example but I would assume that Windows is much larger.
Absolutely. Measurements like that are meaningless in light of the ubiquity of MS-Windows.
The point of the article is valid though, in that you're seeing a lot more companies that were leery of Linux becoming adopters, starting with back-end processing. At my company, they are investing heavily in Linux, and interestingly enough, it's largely on the MS side of the house. They are using VMWare and running multiple instances of Windows on them for those services that are currently residing on single-tasked boxes. Linux runs directly on the hardware, then MS-Windows runs above it. It makes management of MS-Windows much easier.
We'll be doing essentially the same thing on the Unix side of the house when we start a big datacenter move in a couple of months.
You're not far off. MS couldn't grow by 12% without being on every computer that exists, and some that don't.