Predicting innovation is hard. Precise numbers about growth or in the example sited, subscribers, are guesses based on little or no good information. Predicting innovation is a trends game not an exact science. The trends are what show you where the future might end up. Doing any kind of cost benefit analysis on any unrealized innovation is just futile. What you need to do is place bets on what trends will show up. That is the only way to really predict innovation — having enough bets covered to catch the next trend.