With limited resources (as all companies have), we concentrate on those projects that both develop the most cash flow and affect the most number of customers for future growth. What also affects these prioritizations is that there usually is a time frame that technically is optimal (ie: working on some aspect of the application, so might as well add that cool other feature to it while I have the code opened up). We are also skewed in priority when we have a customer who is paying for prioritization (which many of our dealers do, and what the rest of our customers also benefit in - iPort is a good example here).
PocketPCs are cool, UMPCs are cooler for HA / HT control. UMPCs can releverage almost all of the rest of product suite.
Yes, we do know that we should carve up time to work on PPC apps. Just can't promise on when as it's not a front burner.