While it is obvious that the Yield to Admit Ratio is not a direct measure of prestigiousness (much less Prestigiosity, on which I believe @hunt has a copyright), it is actually pretty remarkable how just using (1) the number of students applying, (2) the number admitted, and (3) the number enrolled, one can produce a relatively accurate representation of the USNews and other ranking systems.
p.s. a log-transform of the Ratio will maintain the rank order but give a value that is less extreme at the top ranks.