ScottCC makes some good points. Before MS released the first OS labled Pocket PC, they did a huge market research project. They talked to people who were using WindowsCE devices. They talked to people who were using Palm devices. They even talked to people who had bought and returned said devices. The result was a pretty decent user interface for the time. MS needs to open the channels _very_ wide to user input. Of course they'll get a lot of "It doesn't work." without the "how", but they'd also learn a thing or two about how their products are really used and what would make us jump to upgrade or buy new.
Along with the ideas previously suggested, I'd add a few really wild ones:
* How about some way to look at the WM device as your _primary_ PIM point as opposed to an adjunct to your desktop? Major changes in ActiveSync/Windows Mobile Device Center would be required, especially doing away with the "erase everything on the handheld when you stop synching it" polixy.
* See my
"Synch-less" thread for a really odd/wild idea.
* Buy out DataViz or just license their "Documents to Go" so synching docs doesn't lose formatting and so you can read native desktop formats.
* Through it all, keep in mind that customers buy these devices to have quick, easy access to their data on the go. To that end, make sure you don't lose instant-on operation and streamline the user interface for faster (and less confusing) navigation. (As mentioned above, watch some real users to find out where the bottlenecks are.)