It seems that this would be a much better way of funding research in a general sense, since that seems to be the general process of their problem solving. The only thing is that to apply this to other fields, the barrier for getting people that are equally motivated and intelligent about analyzing if the research is 'going' somewhere is high
