I did this for the first time a few years ago, and it took me about eighteen hours to figure out all of the forms and terminology. And I'm not even running a business, or anything extra-ordinarily complicated. So possible, but time-consuming.
"There was just a paper released about a widely used pesticide that did not appear to actually increase yields."
When my father was in a university geology department in an Agriculturally intensive midwestern state in the 80's, he worked on a few projects with farmers related to water use. One thing he looked into (unrelated to the main research project) was the cost/benefit of ferilizer use. The gist of it was that he came up a curve representing the marginal value added of fertilizer, which started high, and eventually went negative. Past a certain point, less fertilizer was absorbed into the soil and more went directly to runoff/there was more than the plants could use, something like that, and thereafter the added cost outweighed the further crop gains. Most of the farmers in the area were going beyond that point to some extent, and could increase their profit by decreasing fertilizer use to some extent. He said the farmers he talked to were skeptical, but one tried it out on one of their fields that year, found it to be an improvement, and a number of farmers in the area reduced their fertilizer to some extent using the curve as a guidance. The fertilizer sellers would probably just have said that more fertilizers would lead lead to more productivity, which was true, but not presented the cost/benefit.
I don't know too much about farming/crop science, but could see there having been improvements in things like this as well as GMO's & selective breeding. At one point Iowa State University was one of the leading statistics department in the US, which I believe was in large part due to the applicibility of statistics to farming.
"Or all the least efficient farmers were bankrupted out of the market."
One of the trends in the Midwest over the last few decades has been the transition from a larger number of smaller farms to a smaller number of very large farms. Perhaps this is greater efficiency/mechanization/scale.
red = choleric
blue = melancholic
green = phlemetic
yellow = sanguine