Agreed, but to bean’s rather curt/arbitrary point, every analysis has to be scoped in order to be relevant. There hasn’t been more comprehensive analysis than MLR’s, but even MLR’s is limited in scope and still so broad it’s really hard to come to any sort of conclusion.
do a comparative to scope, evaluate, set a grade/standard, do another comparative evaluate, wash, rinse, repeat to develop a kpi dashboard and by the time you are done, Falvey was fired 5 years ago and data sets are stale.