Tom Hollander doesn't think so.  Admittedly, Tom is part of the p&p group at Microsoft, so his opinion might be biased.

Every couple of years, the tech media giants get together to hold a secret meeting. According to my sources, it takes place in a cave, deep underground, somewhere just east of Wagga Wagga. The main order of business is to choose the next technology that will be the death of Microsoft... this time for certain. Past appointees, in chronological order, include Netscape, Java, Linux and Google. And while all of these technologies have proved successful (by at least some measures), none have exactly sent Microsoft packing.

Basically his complaints center around quality and features.

The first one I looked at was Google Spreadsheets, which was simultaneously one of the more impressive AJAX applications I've seen, and the worst spreadsheet application I've ever used. Yes it's cool that some Javascript rocket scientists could build such a thing, but Microsoft Works for DOS had more functionality circa 1988. The word processing offerings were even worse, with most offering less functionality than Wordpad.

And he's correct, for the most part.  And, if you're a professional user, you're probably not going to be happy with the current versions of these new web-based office products.  OpenOffice 2.0 is pretty good, though.  I've used it and have been impressed.  But, here is where Microsoft always misses the boat.  My father-in-law doesn't need ever feature in Word (frankly, neither do I).  What he needs can pretty much be accomplished by Wordpad.  So something like this might work for him.  I think that's the target market for web-based office products.