I'd say that this author is a wee bit biased (of course, I am too). But I don't think he is being entirely honest here. He talks of the perceived benefits of Linux vs what he describes as the actual benefits of Windows. Howsoever, he does have a point on the amount of units sold. Given that, it is easy to assume that windows rules on the Internet, when that (from what I've read) is just plain not the case.
The legal shadow hanging over Linux is a major factor imho. When this is cleared up, or at least there is a bit more visibility, I think alot of corporate IT managers who are on the sidelines will dip the proverbial toe into Linux (or to come out of the closet on their current deployments). The enterprise service and support products from Red Hat and SUSE are really quite compelling, at least server side. Seriously, how many people do you know who are truly happy with say MS IIS or Exchange?
I think the author is being quite realistic. I manage around 45 different servers approx 10% are Linux based and strangely these boxes give us the most grief. We have had a succession of Linux Gurus come and go and still these boxes don't work quite right.
Our Solaris and Windows boxes, on the other hand, never have or rarely have any issues. It's important to note that most of our management tools are Windows Centric (which of course helps alot). But overall management wise the Windows boxes are really quite reliable. (yes even NT4 hehe).
Linux, to me at least, just isn't ready for prime time. I had a fantastic Solaris contractor bluntly tell me that Red Hat and Linux in general was not advisable in a production environment and at this point I tend to agree.
The bias is in asserting that squeezing more money out of captive customers means anything. Ask a Computer Associates shareholder if that is a resilient strategy with a future.