In what probably comes as a surprise only to people who haven't been following trends in open source recently, the vast majority of people who've grabbed OpenOffice.org 3.0 in its official release are Windows users. Out of 3-plus million downloads in the first week or so, almost 2.5 million of those were for the Win32 edition of OO.o.
The most recent stats are even more impressive -- 5.2 million downloads, with more than 4.4 million of those for Windows. That's not too shabby for what is intended to be a mainstream productivity application that doesn't even have an ad campaign. Then again, as my friends in the indie-publishing and -recording spaces have found out time and again, it's word of mouth that's the best advertising of all.
A couple of things are striking, apart from the mere fact that so many of the downloads are Windows. People who download a new version of an app within the first few days of its gold/RTM version release are almost certainly existing users, so the numbers we're seeing are probably a good reflection of how many really devoted, core OO.o users there are.
Keep in mind also that none of these numbers even include the people who get and use OO.o regularly from their local Linux repository. I wonder if that's at least part of the reason why Linux has a smaller slice of the pie; they don't have to go to the OO.o site to get the program, and so their usage isn't tracked this way. It also doesn't track people who use one of the OO.o derivatives -- Symphony, or StarOffice, or Go-OO. This means the active pool of OO.o users is far larger than 5 million -- but, again, the exact numbers are by definition impossible to nail down.
Now, one of the easy formulas I've heard bandied around is that each copy of OO.o downloaded represents that many less copies of Microsoft Office sold. I don't think that's automatically true, though. In my own case, I still have Microsoft Word 2007 and Outlook 2007 running full-time, so OO.o isn't a complete replacement for Office -- yet.
What I think it does represent, though, is a growing awareness of the alternatives and how they work. It is very difficult to unseat any entrenched piece of software, because you're also attempting to unseat entrenched work habits along with them. Most of the resistance people manifest toward changing anything PC-related isn't wholly logical; I've known folks who didn't bother with Ubuntu (and by extension Linux) because -- drumroll -- the default wallpaper was ugly.
I suspect getting people to accept the existence of open alternatives to proprietary defaults isn't meant to be easy. It's how you prove you can not just compete but excel. If OO.o can stand alongside Word the way WordPerfect, WordStar, XYWrite, and, yes, Word itself all used to stand side-by-side as choices -- and do that without needing a major ad campaign or dirty-tricks ISO standards sleaze to "win" -- it will be nothing short of stellar.