Nick Carr explains why Web Office suites will have a slow takeup:
Whatever the flaws of Microsoft Office, most end users are comfortable with it - and they have little motivation to overturn the apple cart. What is absolutely unacceptable to them is to take a step backward in functionality - which is exactly what would be required to make the leap to web PPAs today. Web apps not only disappear when you lose an internet connection, they are also less responsive for many common tasks, don't handle existing Office files very well, have deficiencies in printing (never underestimate the importance of hard copy in business), and have fewer features (Microsoft Office of course has way too many, but - here's the rub - different people value different ones). Moreover, many of the current web apps are standalone apps and thus represent an unwelcome retreat to the fragmented world of Office 1.0. Finally, the apps are immature and may change dramatically or even disappear tomorrow - not a strong selling point for the corporate market.
There's another thing about connectivity, and it's going to be true for a long while yet: travel. Say I fly to Europe, or Asia. I might well want to do some work on the way. If I rely on "Web 2.0" apps, I'm stuffed: there's no connectivity in the air, and there's not likely to be any anytime soon. There is power though, on an awful lot of carriers. Which means that I can work if my tools reside on my laptop.
If they reside in the cloud? Not so much. Until they get that teleportation thing working, expect to see Office on the laptops of the traveling business guy.