If Outlook suddenly became twice as efficient, you'd expect to see "Time Spent" decreasing.
That said, I think determining the utility of a product from the amount of time spent in it is a very difficult thing to do. Deriving utility from time spent, and then comparing it between products seems very nebulous at best.
Personally, I'm fine with RescueTime's presentation of the graph here. In the end, eyeballs are eyeballs. I don't know if you could generate a metric (from RescueTime, anyway) that could measure product innovation.