“Most cited” is a very poor way of measuring quality and is massively subject to gaming. When it comes to review it’s not uncommon to get comments like “you didn’t cite this paper” where said paper is only marginally related so it’s clear that the reviewer has some interest in you citing it.
'publish or perish' model of academic career progression is a plausible explanation for the general slow down of scientific progress, as scientists are nudged or driven toward research which have a good chance of being published. Breakthroughs, which always come at the margins, is 'low ROI' and therefore only pursued by enthusiasts who are prepared to risk their futures or have the wherewithal to absorb the L's. I wonder if this applies also the Chinese scientific research, probably does, PRC is highly driven by KPI's