Science works that way in the minds of scientists, whose vast experience gives them dozens of actual data studies they've read to speculate from. Researchers have read about thousands of cases, they're the ones with the refined intuition, not forum readers.
Specifically: Is it productive for HN users to have anecdotal evidence rise to the top, or actual peer-reviewed evidence from studies with N=100 or N=10,000? I'd argue if layman HN users have time to read only one comment, we should upvote the N>1 comments, not N=1.
From the guidelines: "Comments should get more thoughtful and substantive, not less, as a topic gets more divisive."
We went from an international article (N=745k) to anecdata, which is a downgrade of substance in my view.
Somewhere in the last couple of decades, the SCIENCE WORKS, BITCHES people seem to have forgotten how real science is actually done. Intuition, anecdote, common sense, hunches are very important parts of the process.