What if many people are willing to pay you a lot, but those people don't exist?

I'm a little confused by this - I'm not sure what to make of the word 'many' when applied to non-existent people. Do you mean potential future people?

Yes, I meant "do not exist yet".

If we could reliably tell these apart in advance we would no doubt make significantly faster progress.

We have heuristics, and they help. We now know that funding basic research that investigates the nature of the world, but doesn't provide any tangible benefit is worthwhile in the average case. I believe that one of the many factors in the acceleration of increases in technological development is this principle, and as fundamental science funding has increased, we have after-the-fact discovered many important things we ended up needing to know, with much less lag time than decades or centuries ago.

I certainly can't attribute all that to funding research that has no immediate application, but that heuristic has increased our rate of advancement.