It's definitely an uphill battle...
...but it has a couple of things going for it. First, XForms is a long long way from ever becoming ubiquitous. Of course, MS could change their minds at a moments notice but at this point IE seems to be a dead end product from their perspective. MS innovation in their browser was a casualty of the loss of competition.
Second, MS is going down the path of Avalon and .Net, while Sun and IBM are busy with Java and J2EE. Both pretty good technologies for what they do, but neither of them has made significant strides on the client side for internet based applications. Java applets have proven too clunky and MS has a history of trust failures with trusted ActiveX components. Neither technology is a sure bet. Java's pretty set as the Enterprise glue, and .Net may succeed for intranet based apps. But when it comes to over the internet apps, both technologies still are just cranking out html based forms and doing a lot of processing on the server.
Third, there is a pent up demand for improving the forms processing capabilities of html. While DHTML, CSS, DOM and JavaScript have seen a number of improvements over the last few years, web forms are still stuck in the rut. Lot's of web apps could definitely use the enhanced forms processing capability (I know several of mine sure could).
Finally, the proposals for extending the HTML forms capability is the only solution presented that represents an incremental solution to rich internet client apps. It may not be the most elegant - XForms being much more well thought out, having started with a clean slate. But it is incrementalism that usually succeeds in such matters. Most don't won't to learn a completely new way of programming, they'd rather just extend what they currently have.
That said, the odds are probably against this rebellion. But I think that the odds that this will succeed are about as good as any other internet based proposals. IOW, the possibility of failure doesn't mean that someone else wins. It may just mean we continue to be stuck with HTML3.2 forms based processing out to the indefinite future.
