In science, we have independent peer review, and if we have someone who's truly independent from the process, they'll be able to weigh up the different datasets and try to come to a conclusion, recognizing that uncertainty, and with some transparency of that uncertainty.
The problem with the CSAS process, I believe, was that because the panel was dominated by one type of person with links to the industry, the consensus didn't reflect the complexities in that data and the different findings. Within that consensus process, some of the important work that had been done and those conclusions found...those ideas were suppressed.