It's got NOTHING to do with the OS at all - stop the myth
The reason the tablets didn't take off and the iPad is is not just marketing, it's simply technological advance. I have used tablet PCs before (usually laptops with a bastardised screen hinge) and they were not usable, because - OK, big insight, ready for it? - they were FAR too cumbersome and heavy. And call me silly, but I rather like to have the screen write where I put down the pen, something I never managed to achieve even after repeated calibration.
The technology to make a usable screen has arrived in bits and pieces. Apple (I think) pioneered the multi finger interface, and at the same time the screen themselves got better, and battery technology improved to a point where you could actually make a tablet PC that WAS indeed a tablet in the physical instead of the biblical sense, i.e. not the weight of a large stone.
However - all of that is hardware. Once you have a decent platform to work on, only then software becomes important, and I have yet to see anyone taking advantage of a decent working tablet I/O. First off, the Apple stuff is operated by fingers which doesn't make for large precision (thus avoiding calibration issues), but does enable multiple ways to provide input..
With a tablet you typically use a pen for input, and that requires some precision. It also requires good recognition, and if I look at the handwriting of most doctors I know where the challenge lies..
But before you go near this, answer question one: what would I want it for? The clever stuff about the iPad marketing was that is gave some possible end user uses. Without having a decent use there is no market, without the market there will be no investment.
Personally, I see tablets and book readers combine, at which point this market WILL get interesting. But I still cannot see any difference between MS, Apple or Linux on the device other than as a factor that determines cost, compatibility and comfort.
From personal experience I think using MS will mean it's going to look good in the brochure and be cheap to buy and you'll bleed forever after (as MS has stopped selling anything but badly finished beta software since W2K).
From Apple it will look fantastic, will probably be quite usable but expensive, and only on one type of Apple made hardware (which is why it will work as long as it doesn't need an antenna, which is clearly not their expertise).
With Linux it will work, but you'll have to choose from 20 distros and it will never quite work completely or support *all* your hardware, and any effort to find support or help is met with the comment that you can find it all on the Internet. Which is true, after long searching you'll find a HOWTO that for the average end user might as well have been written in Swahili and involves using a command line and editing text files with vi.
Let's get some hardware before we start looking at software. Apple has produced the first usable version (although I think it's too think and it MUST have a battery you can replace), and I think the first "open" platform will be found in the electronic book sphere. From then on the fun will start.