The data in the first table is from the 1991 cohort (class of 2013). Data from the third table is from the 1980-82 cohort/college classes from about 2002-04. Obviously these data are old, and I’m personally not comfortable with their utility anymore.
Some schools have improved since then, some others might have not. The data I shared above from the 2021-22 CDS show the same percentage of full pay families currently at each school (63%).
Beyond that, I don’t think rankings dictate what is better for a given student. At least, they shouldn’t.