It's a good question. We've tried to answer it in the paper.

tl;dr

Georgem is basically right. It's the mother ship, like it or not. If it shows signs of failing now, we've all got a big mountain to climb from the negative press.

Secondly, it turns out there is a pretty decent business model in the approach we have taken that can support an incentives program.

This happened to me a few times before. You create a business plan to do one thing, but once you look deeper, you find a little golden nugget you were never expecting to find.

Anyway, that little golden nugget should be good enough to support many thousands of Bitcoin nodes. You would only get thousands of Bitcoin nodes earning a profit, if they are running on thousands ServiceNodes.