The problem is that the incentive is biased against scepticism. So the process is more likely to find potential pathways but not notice obstacles or counter arguments.