Accepting the premise, I think the most interesting question is: how do non-farmers learn enough about farmers' problems to solve one?
Generally, how do non-specialists find enough out about a specialized world to solve a problem from that world? And then, learn enough about the specialist culture to successfully sell into that culture?
Short answer is: they're brought in as "consultants."
By "consultants" I don't mean blue-suited guys from Accenture. A consultant could be the local machinist who is asked to make a gearbox to connect two diesel engines together and then realizes that he could sell a bunch of them to other farmers.
There is considerable value in having friends whose interests are radically different from your own.
You have to spend significant time there, would be one way. A related but seemingly tangential question is when a young rancher is about to be married, he would do well to pick a rancher's daughter.
Better for the farm boy to go to college, pick up a wide range of ideas, go back to the farm. Probably the best way to get outside ideas implemented.
Right -- the big question here is, What is the optimal amount of knowledge you should have about a problem / market? There are diminishing returns at some point.
And I don't see "diminishing returns" in knowledge as significant. There is always something else relevant to learn, maybe in another field but still relevant.
Generally, how do non-specialists find enough out about a specialized world to solve a problem from that world? And then, learn enough about the specialist culture to successfully sell into that culture?
(Not accepting the premise, a few clicks through gives its origin: http://en.wikipedia.org/wiki/Jane_Jacobs#The_Economy_of_Citi...)