Well, actually "it depends on what your definition of 'is' is" -- Bill Clinton
If the argument is that "sometime in the future this functionality may exist, and therefore it is 'capable of this functionality'" then there's no possible way to argue against it other than saying, it doesnt exist currently. Something major when dealing with linux is hardware support. You either have to wait for the manufacturer of hardware to produce a driver, or wait for some other shmuck to write one that will probably crash a lot and support 10% of the functionality

Over time for specific products the natural progression is to improve the drivers that are open source, sure.. but... time is a key factor. I don't buy a video card to have it work next year, I want 100% functionality RIGHT NOW (video card isn't really a good example since I believe ATi and NVIDIA both produce linux drivers, but to illustrate when I want the usage out of my hardware). Something else is that it is *difficult* to burn a cd in linux. CD burners bought off the shelf at a store come with cd burning software like Nero Burning ROM. Install it in windows and you're good to go. Linux on the other hand you might have to update your kernel (while talking to someone about cd burning they mentioned that a short time ago you had to emulate SCSI to burn cd's), download and try 75 open source cd burning solutions, and might even find that they dont work right on your burner. I have a friend who has used linux as a desktop system for years, and he had the worst time trying to get his to work. Anyway, in short term without expecting too much from the operating system and software selection, I'm sure you'll be fairly happy. Long term once you find stuff that you absolutely need to do at some point in time and simply cant without a big hassle, thats what really counts. Just about everyone is pretty happy with what they get right off the bat, and it takes some time to discover what you need.
Open source is actually one of the worst possible things. It sounds good on paper, but in reality 95% of open source products suck ass. MacroQuest and MQ2 just happen to be part of the 5%. For example... do you want an open source StarCraft, which was developed in gas station attendants spare time (note: I have no idea the jobs of the people who work on freecraft specifically), or one that a company has actually invested money in hiring developers and artists for and is developing with profit in mind? One sucks, one is an awesome game that sold millions of copies for good reason. I'm not saying there aren't good software packages for linux, but most of them that dont suck ass are server applications. Something I use on windows every day is open source and made for linux -- X-Chat ... unlike most linux apps I find it superior to other alternatives (mIRC). Just one problem... whenever I get disconnected and it starts reconnecting, it crashes after a few tries! such is nature i guess. And yes you can say "gaim" or "openoffice" and be popular with the other open source nuts

and that stuff like apache is open source (but thats a server so its not going to help the case for desktop system)
I vehemently disagree with "more innovative software", and in most cases it's actually LESS innovative software (note how long it takes for OpenOffice to catch up to MS Office). Sharing of ideas is what leads to innovation, not necessarily sharing of the source. Even then it actually depends on
who is interested in innovating on that project -- maybe its lots of people, maybe its just some guy who wants to as a hobby. Another thing is these people need to eat, meaning they need an actual job to pay for their habit of writing open source software :) Lots of them will be software developers that are working at some company, researching, innovating and developing software for profit. Sometimes companies take interest in developing open source, but it's hard to make a profit when you're just giving stuff away so not many put real time into it (some might add something to apache to run on their web server or submit a bug fix, but does that really count? and yes I know that selling "support" is the key, thats why they make all linux apps difficult to use!). Anyway, the stuff theyre researching and creating at their place of employment is going to take precedence because theyre getting paid for it, and the open source "on the side" innovations are going to take much longer and be much further between.
I'm going to argue that real innovations usually come from universities doing public research, or research groups in companies... which is why when I'm looking for innovations to implement I will search for public research papers :) A lot of that research is tied into small open source solutions for proof of concept though.
Anyway, I'll buy the learning thing.
Note: light S&M is ok, my wife enjoys being tied up a bit

now.. actually scarring someone, drawing blood or having bruises... thats just fucked up