[_delirium]

I would say that closeness in time to commercialization is more important than bottlenecks per se: doing research that will enable great products in 20 years is rarely lucrative, because it's difficult to capture any of that future value (especially if it's further out than the length of patents, and often even if it's within that length).

So it's smarter (if you want to make a profit) to let someone else do that, e.g. someone who's paid as a researcher and isn't trying to turn a profit, and instead look for things that are 1-3 years out. You even see it within academia; applied math pays a lot more than pure math, for example, even though both are quite important to mathematical progress.

[001sky]

An interesting example of your case: the Art dealer. Despite doing none of the work of art, they collect 50% of the sale. I think the missing link here is that the effort involve in innovation is S-shaped. [1] Like in your idea of original research, The original idea or creative spark, overcoming inertia. etc. Next, there is a lot of stuff to do to get it roughed out, but many competent people that would not have the genius to originate the idea, can/do help move it along. But then their is finishing and integration, and again for this you need a master (eg, steve jobs). Likewise, with a business. In business, the last step is sometimes looking like the art dealer. A bottleneck? Yes. But being an art-dealer is still a bit of work. They have to make it marketable. They have to market it. There is a needle in a haystack problem. There is knowing about the haystack (rolodex), doing the legwork to meet clients. Then running the shows, exhibiting, writing about the "value" and context of the work. Etc. The pay (like you say) is higher than what seems reasonable. (Although, perhaps not as easy as it seems.)

[1] http://upload.wikimedia.org/wikipedia/commons/thumb/8/88/Log...