Wheat still required far less work to domesticate. If you compare the ancestor of wheat to the ancestor of corn, it's clear that corn changed significantly in the domestication process.
Presumably, though, at every stage it's something worth eating (at least by the standards of what else is available) - nobody's going to start farming something inedible in the hope that their grandchildren will make the first bread from it.
Ultimately most of history is also conjecture. But conjecture can be quite exceptionally accurate given enough reference points to compare against. For example say you have two manuscripts: A and B both copying a certain document. Manuscript A says eorum (of his/its [pl.]) while Manuscript B says earum (of her [pl.]). With just these two manuscripts we have no way of knowing which is "correct"; what the original said. However say we found Manuscripts CDEFG and manuscripts CDE and F have eorum while only manuscript G also has earum. Given the larger body of evidence it seems far more likely now that the original text had eorum and B and G were either mistakes in copying or an attempt by the copyists of B and G to correct a perceived mistake in their source texts. Now - we have no way of knowing for sure that ACDE and F are correct while B and G are incorrect (in representing an authentic reproduction of the source text). For all we know eorum is an error in A which was incorrectly copied into CDEF, but given the evidence we have at hand, the supposition that eorum is correct is the most likely scenario. Subject to change, granted, but the most likely postulate we have given current evidence.
This isn't to say history, or archaeology, or manuscript analysis are irrefutably correct and representative of the absolute truth. The fact that historiography exists is a testament to that. However there is very very little (if anything at all) which is irrefutably correct in this world. Distance (in time) from the event in question only weakens our ability to approach the truth in that the number of data points likely to have survived for us to read and interpret now is smaller. Just because topic A has a surviving written chronicle while topic B has no recorded primary sources (but an absolute butt-ton of archaeological findings) doesn't mean that topic A is in ipsi more factual, reputable, or accurate.
Interesting explanation - though (if my single-data-point conjecture that you've started a course involving Latin is correct) ipse is ipso in the ablative, not ipsi. As in manuscripts, though, I would give more weight to the reliability (or not) of sources in working out the truth than number alone. I probably wouldn't set much in store by a thousand North Korean films about Kim Il Sung.
Also intrigued by:
Distance (in time) from the event in question only weakens our ability to approach the truth in that the number of data points likely to have survived for us to read and interpret now is smaller.
I think it works the other way too, at least for a while, because people involved in the moment inevitably have only a partial picture - they see great big historical events and processes as only what they can literally see with their own experience. We can see this even today, with a lot of recent history being re-written in the light of new discoveries and declassified documents. A good example is people talking about Operation Yewtree etc and linking it to the workings of politics, suggesting that some people were promoted into senior government positions because the PM had leverage over them. Whether or not it's true (though I'm not sure that interpretations like that can be simply true or false), nobody could have joined the dots in that way at the time, or even five years ago.
I'm not sure 'the truth' with a lot of the interesting questions in history really exists, in the sense that if you could only get all the data together then you would have an 'answer', and you could prove to anyone else that it was the only answer. That may be true for questions that we used to answer in school, like 'when was the Battle of Waterloo?' or 'who was king after Victoria's death?', but I don't think it holds for actually interesting questions - questions like 'why did the First World War begin?' or 'how democratic was Britain in 1913?', or even 'what was it like to be a Roman slave?'. Just as the data is only ever individual perspectives on things, which are often less 'better' and 'worse' and more 'different', I'm not convinced that historical interpretation can ever be (past a certain point) anything other than 'different' and 'more convincing' or 'less convincing'.