It seems to me that the potential success of the realtime API approach would be reliant on impressive, profitable applications. Although I can think of a few, I

It seems to me that the potential success of the realtime API approach would be reliant on impressive, profitable applications. Although I can think of a few, I can't think of enough of them to argue against the AdWords approach, which apparently drives revenue for now. If Twitter had chosen the API direction initially it would have been partially on them to cultivate the developer community required to make their strategy works — which could be seen as orthogonal to developing a great product and solving infrastructure problems. In the long run these two approaches may not actually be at odds with one another. AdWords-ish works for now, and as new apps are developed there may be an opportunity to move back in that direction.