You can read his review here.
After six years of working on desktop Linux full time myself, unfortunately, I still have to agree with Walt's conclusion: "But for now, I still advise mainstream, nontechnical users to avoid Linux."
Ubuntu and other Linux companies are doing a lot of good work, but it's a big hill to climb. As much as people may dislike Microsoft and Apple, they have an enormous head start on hardware and software compatibility. It's tough to catch up. It's happening, but on the desktop, its still a slow, long and arduous journey.
However, as big of a technical job as it is, the REAL challenge is more with the mentality and dynamic of open source development. This is why articles like Walt's are GOOD for desktop Linux, because they helps the open source Linux community to understand that the average person has a very different definition of "ease of use" than a technical user does. During the last several years, desktop Linux has made big strides in basic functionality, but as Walt points out, it will take that next layer of polish to be ready for the mainstream. (By the way, here is what a mainstream market looks like.)
The REAL challenge is that desktop Linux was developed by, well, developers. Once the open source community has scratched THEIR itch, they often turn to other projects. It will take leadership from companies, such as Canonical, Novell and Red Hat, to keep the development going to scratch the itches of the mainstream nontechnical user, which are very different than those of developers.
There will be enough inherent incentive for the open source community to get Linux "to work," but it will take additional incentives to get that same community to continue making it easy for everyone. That incentive will be, as it is with most things, money and the power of a free market. It's no cooincidence that the most popular Linux distributions are from companies with serious financial dollars behind them (Canonical, Red Hat, Novell).
Will Desktop Linux ever be ready for "mainstream, nontechnical users?" I honestly don't know, but if it does happen, I predict it won't be for at least another five years, and it will come from the commercial companies with incentives to take it to that next level.
PS: You can read more on my views of the free market and open source software in this previous blog.