My default assumption is that "don't try to engineer things to be better" is the wrong answer. If we don't understand it well enough yet, just experiment more until we do. The awesome thing about indoor farming at scale is that it becomes really easy to do lots of experimenting and zero in on exactly what things each plant requires to thrive.
So let's do the opposite thing. Get serious about industrialising farming. No more below-minimum-wage labour, no more flooding pesticides that drain into rivers, no more guessing about what happens because the weather is random. Figure out how to do it cheaply, space-efficiently, and near to the demand so that we aren't driving trucks all over the place (burning a lot of petrol) to ship from middle-of-nowhere farms to urban centres. Just because we haven't got solutions to all these problems today shouldn't stop us from trying to find those solutions.
So let's do the opposite thing. Get serious about industrialising farming. No more below-minimum-wage labour, no more flooding pesticides that drain into rivers, no more guessing about what happens because the weather is random. Figure out how to do it cheaply, space-efficiently, and near to the demand so that we aren't driving trucks all over the place (burning a lot of petrol) to ship from middle-of-nowhere farms to urban centres. Just because we haven't got solutions to all these problems today shouldn't stop us from trying to find those solutions.