The numbers are interesting and though they don't look pretty, I'd say that a selection-bias makes the average daily/sales number (from the sample of 94 paid apps) a lot higher than the real average-sales number for the 100K+ paid apps in the app-store.