MS fights back
The next step in this chess match is that MS won't take this lying down. On the Office front I don't think they can do much to fight back. Office applications should be a commodity -- they ran out of good new features years ago. ("Ribbon", anyone?)
Tackling the Windows beast is significantly harder. First, there's a question of how it can best be done. As mentioned, WINE running on Linux accomplishes this to some extent, but it has to be great. This is where I apply my "soccer mom test": a totally stressed-out, novice PC user soccer mom must be able to use the system easily, or it won't work. IMHO, even Ubuntu is not soccer-mom-ready yet.
The hurdle here is that MS will do anything they can to create FUD (fear, uncertainty, doubt) in the minds of end users ("Gosh, will my XZY application work on the GoogleOS, or should I just go with Windows???"), and Google must do everything possible to eliminate this. If "free" opens the minds of IT managers and consumers, other things like a "GoogleOS Certified" campaign must be established to eliminate the FUD.
Assuming Google does succeed, MS will eventually be selling two products that others are giving away for free, and assuming they are on equal footing, that's not a winning hand.
Execution, the other strategy
As I mentioned earlier, the other strategy for Google is to ignore MS and simply keep executing their vision better than MS. Google currently has great technology, and also seems to have great development velocity, constantly adding new features to their ad serving technology, so you can say that if this is a race, Google is running it faster, and pulling away from MS.
Assuming that Google has a ton of additional ideas already waiting to be developed (that will enable them to continue widening their technology lead on MS), creating their own versions of Office and Windows may be a distraction. But, if they don't have a stable of new ideas, that points to the search and ad markets slowing down, and moving more towards "commodity" status themselves. Instead of Google running away from MS, MS will be catching up with them.
Microsoft is attacking Google's search and ad bases, and I believe that they will keep attacking Google as long as they have other strong revenue bases to lean on. From Google's perspective there are several strategies for dealing with this threat. One strategy is to keep executing their current web-based vision, including their search technology and Google Apps, hoping to execute better than Microsoft (hoping the MS will eventually give up, or be so small in the market as to be a non-threat).
Another viable strategy is for Google to in turn attack the Microsoft "cash cows" (Windows, Office) as a means of protecting their own business. By gutting these MS revenue sources Google can make it much more difficult for MS to pursue their search and ad strategies.
In creating their own GoogleOS and Google Office applications Google can build on open source efforts such as Linux, WINE, and OpenOffice (alternatively buying Corel Office). To succeed, the learning curve for these systems must be very small, so they'll pass the "soccer mom" test. Google Office must be fully compatible with MS Office, and GoogleOS must be binary-compatible with MS Windows, and initially a list of "GoogleOS Certified" applications must be created.
The following references were used in creating this article.