I find this an interesting take on the subject. I'd use the same line of thought to argue for the opposite position: it's a good thing if a statement could be proven to be wrong! Arguments that rely on a set of empirical claims of superiority are good arguments, because they can be tested. Statements that are not refutable -- such as statements about the superiority of a subjective moral -- are bad arguments.
For the record: Here's one Open Source advocate who applauds LinuxCare for saving themselves by abandoning Open Source. I guess, the current users of LinuxCare's products probably agree. Yes, I advocate for "proprietary development methodologies in areas where evidence seems to show that they are more effective".
An idea is not inevitably good just because it's guided by principle.
In fact, I believe most ideas that are guided by principles -- especially those that claim moral superiority of a certain principle -- seem to turned always into human tragedies.
Summary: there are two camps of open source advocacy: those who state their position on matters of principle, and those who make their case based on promoting measurable quantities such as higher quality or more flexible software etc. Author believes the former is better because "Principles are based on a type of Utopianism; they are a statement of how we think things should be" whereas the latter argument is open to obvious, demonstrable cases where it is false.
While I have every sympathy for his intellectual position, as somebody who has had to make a business case for using open source software, I have found the vocal "efficiency" camp and its publicity invaluable when dealing with an audience that is sceptical towards matters of principles.... on principle. Yes, the superiority argument isn't always technically correct, but is certainly more appealing to certain classes of people.
I've always found there to be two primary camps as well, with the following division.
The first group is the Stallman-esque group. They believe that people should use software because it is free (free as in libre, of course, not gratis). Being good is secondary; the fact that something is free should be sufficient for people to prefer it over other solutions. This has potential negative consequences for developers; a developer who believes in this may be less motivated to make his program better than proprietary equivalents because he does not believe that it needs to be so for him to promote his program as an alternative. The Stallman-esque group also tends to believe that marketing is more important than engineering; open source is about advocacy more than it is about coding.
The second group is the Linus-esque group. These people believe that the primary goal of open source is to produce, via a community process, really good software; freedom is secondary. That software will become popular not because of marketing and advocacy, but merely because it is so good that its engineering merit will stand on its own. Users should use what software is best for them--it's the developers' job to ensure that "best" means "free". These people are often seen as "impure" by the Stallman-esque group because when asked for the best way to set up a system, they will often include non-free software in the mix--they would design the system for maximum effectiveness, not maximum freeness, since their goal is to produce the best system possible, not the most free.
From what I've been able to tell, the divide between these two groups seems practically unbridgeable; the philosophy differs so much that the two sides often completely fail to understand each other, much like the divide between GPL and BSD advocates.
That is what he describes, his point being that rationalisation aside, let's face it we do this because it is fun.
"the divide between these two groups seems practically unbridgeable"
I don't know enough to comment about whether that is accurate; but I would say that this is one divide that doesn't necessarily need bridging. We're not talking about VHS versus Betamax here; as long as both sides keep doing what they are doing for whatever reasons, everybody wins.
But often the sides come into conflict. For example, most recently, the Stallman-esque group started a massive marketing campaign for Theora--it's free, so we should use it. This caused the Linus-esque group incredible aggravation, because they cannot understand why someone would want to promote something that is technically inferior in many ways. Equally, the Stallman-esque group cannot understand why the Linus-esque group would even consider using multimedia formats covered by patents--why would anyone do that?--it's against the open source philosophy--and away the argument goes again.
there may be empirical evidence for your claim but I don't suspect one can easily group programmers into these two groups so readily. As an example Stallman himself has written some pretty good code, for instance emacs.
For the record: Here's one Open Source advocate who applauds LinuxCare for saving themselves by abandoning Open Source. I guess, the current users of LinuxCare's products probably agree. Yes, I advocate for "proprietary development methodologies in areas where evidence seems to show that they are more effective".
An idea is not inevitably good just because it's guided by principle.
In fact, I believe most ideas that are guided by principles -- especially those that claim moral superiority of a certain principle -- seem to turned always into human tragedies.