A new approach to plant breeding

That worried me too. Unless the survivor seeds for each generation were grouped to each locality with conditions of climate, soil and prevailing pests you are not going to get a useful shift in the gene pool as the requirements of any given locality are at risk of counteracting the others. It reads like there is but one breeding pool spread over a wide area in which case there will be little change unless the starting stock's genome had a weakness that all (or most) locations would alter for the better. Why would you start with inferior stock? I shall ask some experts and report back if I have missed something.

David