About a decade ago I was (with my colleague Peter Andras) doing a lot of scientometric research on university research performance (mainly in the sciences) - and (because I wrote regularly for the Oxford Magazine) we looked closely at the performance of Oxford, relative to other British Universities.
To our surprise, we found that Oxford did not stand out from other British universities - except in volume of production. Once we too account of the number of people, the quality of research (insofar as quality could be measured scientometrically - which is not very validly - albeit the direction of bias would tend to favour, not disadvantage, Oxford) didn't seem any different from the other big British research orientated universities.
Take a look for yourself:
In sum - Oxford was bigger - but not detectably 'better'.
(Later on, we found much the same for Harvard. MIT - on the other hand - was outstanding!)
Note: I completely stopped doing scientometrics, mainly because nobody at all was interested. Universities nowadays are run by managers - and managers are not interested in objective research - indeed they are actively hostile: they just want to cherry pick the data to justify doing what they have already decided to do.