I purchased a new bike this year. It came stock with very high-quality ('race' level) 38mm gravel/cross tires, which roll very well indeed on paved surfaces. Hated them; bike felt 'out of focus' for me and the riding I do.
Switched them out to my preferred 32mm, very high-quality road tires: bike instantly came into 'focus' -- for
me.
As an old (68), arthritic, recreational cyclist, that is all that matters,
to me. I would no sooner go back to 38s, or 42s, 2.3" or whatever tires on my bike or another bike than I would go to 23s or 25s: 32s are my sweet-spot. Anything wider makes the bike feel heavy, sluggish, and slow -- for no perceptible increase in comfort --
to me.
However, I would not dream of arguing from my experience/preferences that all cyclists, everywhere, riding on paved surfaces should be riding on 32s and only 32s, let alone that there is some sort of "scientific"* basis for such an idiotic assertion.
But hey ... this is teh Biek Forms, where personal preferences are asserted to be universally-applicable truths all the time.
Carry on.
*"Science", in some recent posts on this thread, appears to consist of some very loose paraphrases of some of Jan Heine's blog posts on tire width.