I'm the editor of CDM and also run our servers. (apparently not terribly well, though I am in the middle of a migration to a new server config -- should've, uh, waited on this story!)
I'd like to get more data on hardware, too, and I'm curious what's been giving you trouble. I regularly see various problems on all three platforms, though I agree Linux is probably the least familiar and needs information dissemination most urgently (at least for music production).
While I'm waiting and restarting Apache (cough), some of the things folks are claiming here seem to be misinformed. That's not necessarily their fault; it illustrates that better documentation is needed, and simply pointing people to audio-centric distros I think is not enough.
For driver support -- RME fills the pro audio gap nicely if you're looking at the high end; they're the ones that are really doing it right. You'll also have good luck with any class-compliant USB audio interface. I'm getting good results out of a Cakewalk SPS-25 (basically equivalent to an Edirol UA-25). FireWire support is greatly improved, and you can check there at ffado.org. Most internal chipsets are also well-supported by ALSA - not a high-end option, true, but it means you can mix something on the road listening to the headphone jack without having to muck about with something like ASIO4ALL on PC.
Software: it's true there isn't much in the way of Linux-*only* software, but not that you don't have choices. Renoise and energyXT now both run natively, Renoise being a huge deal to fans of trackers. And many Windows apps can run better under WINE, with ALSA, WINEASIO, and JACK, than they do on Windows. That's the reason Native Instruments software can run on the MUSE Receptor, specialized hardware that runs Linux and WINE under the hood. It's solid enough that it winds up being preferable for people to buy that hardware over a laptop - yes, even over a Mac laptop. You probably won't get that kind of reliability out of a Linux setup out of the box, really, regardless of distro. But if you can set it up in a way that will be rock-solid, that could be worth the time for people.
I think that's the bottom line: a lot of people would be happy to invest a little extra time and effort to get an open system running. The problem is, they don't know how. The responses here demonstrate that people aren't aware of what some of their choices are, or have had (understandable) frustration because the distros ship out-of-box in a way that doesn't quite work, and there's not much clear documentation to tell you how to fix it.
ALSA isn't perfect, but neither is Core Audio, let alone ASIO. ALSA combined with JACK can be an exceptionally-terrific audio system for these applications.
So, I can make you a deal - I will put more of that information online, *and* fix our servers, too, so you can actually read it. ;)