However, isn’t it more of a numbers game than the quality of PhD students/graduates?
A professor at a research university supervises a few dozen students to PhD completion over his/her career, far more than needed to replace him/her when s/he retires. Obviously, some people with PhDs do go to research/academic jobs other than research universities (e.g. a few go to LACs, more go to non-research universities or community colleges, and some in some fields go into industry or other non-university research). But that still leaves many more PhD graduates than there are research/academic jobs for them, except in a few fields where industry hires the “excess”.