I disagree with this aspect because I think scientists can not and should not be trusted with Bayesian Inference -- mostly because firstly, science should be sure of it's conclusions, rather than assigning probability estimates to different interpretations. The potential for human judgement to be wrong coupled with the fact that often, wrongness has large consequences in the field of science warrants me to believe that the risk is not worth the possible speed of progress that we would gain.

In practice scientists often look at multiple hypotheses at once and consider them with different likelyhoods.. Moreover, the dangers of human judgement exist whether or not one feels one is "certain" or is more willing to admit uncertainty.