I know many disagree with hiding comment scores, but in my view, the quality of discussions has notably improved since scores are hidden.

I would be interested in knowing how pg estimates the effect of this tweak.