In this particular example, from a more global point of view, it would be best if the farmer who discovered new food was incentivized to share the knowledge without restrictions, as widely as possible, because not doing so means lots of unnecessary people sick or dead.
> If people do work, and we benefit from it, and they don't benefit from us benefitting, that's crooked.
I agree, but there has to be diminishing returns on that benefit. Otherwise, each discovery is forever holding the civilization hostage. You can't run an economy based mostly on rewarding the estates of inventors for their past inventions. At some point the discovery has to be owned collectively, by everyone, and become a building block for next discoveries.
Current western IP systems sorta recognize that, at least in theory. In practice, we're dealing with a) protection periods not reflecting the reality of modern industries, and essentially putting a brake on progress; b) a system that's thoroughly gamed, and no longer serves the interests of society. Between ridiculous copyright extensions, vague patents, obvious patents, speculative patents, trolling, rights trading, MAD via patents, the system legitimizes rent seeking, and does not incentivize people to create/discover things in order for them to benefit the whole.
You might argue that "finding" or "discovering" stuff is not work and in naive cases you are right. Regarding wheat, though:
* There is a lot of work that has gone into virtually all edible wheat varieties. We don't eat wild wheat.
* Finding stuff you can actually eat was at one time a lot of work, and sometimes dangerous.
* Saving seeds and saving the knowledge of identifying characteristics of food stuffs was at one time far more difficult than it is now.
If people do work, and we benefit from it, and they don't benefit from us benefitting, that's crooked.