Techdirt has a great response to those that believe that "free" isn't everything in building a next-generation music business:
Of course it isn't, you dolts.
Well, Techdirt tried to be more diplomatic than that, but it does a good job of ripping into those that criticize free distribution of music as a viable business model. The problem with such critiques, as Techdirt notes, is that no one is arguing that "free" is THE business model. As in music and open-source software, the argument is simply that free makes a great component of a killer business model.
Pointing out that it's not enough to give away one's products tells us nothing, especially since many examples given are of lame bands that couldn't make it whatever their business model:
...[T]here will always be some bands that won't be able to make money -- and that's because they're not very good. No business model will work if you're not that good. So, seeing a particular business model not work for some bands is hardly a condemnation of that business model.
This same principle holds true in open source: if your software is junk, no licensing model can save it from being junk. Garbage in, garbage out.
I'm firmly convinced, along with Techdirt, that "free" makes a fantastic component of a larger business model, one that requires a lot of effort (marketing, sales, etc.), a great product, and a lot of luck. "Free" removes barriers to adoption and makes it easier to disrupt established players.
This is why Mark Shuttleworth isn't worried about making a dime on the desktop, at least, not in the way that Microsoft has. Mark made this clear to me earlier this year: free desktop services pave the way to sell other, more interesting cloud services. "Free" isn't the whole business model: it's just one critical part.
In short, don't throw the free baby out with the open-source (or music) bath water. There's tremendous value in giving things away, provided you do it correctly, and I'm not talking about lifestyle businesses around open-source projects like some aspire to run. I'm talking about industry titan-esque businesses.
I'm talking about businesses like Google and Red Hat that I believe will dominate computing for the next 10 to 20 years.