Perhaps not, but the quality of research at IAS is surely better than that at all but a handful of universities. Fair enough, but they also have much "better"

Perhaps not, but the quality of research at IAS is surely better than that at all but a handful of universities. Fair enough, but they also have much "better" researchers than most other universities. My hypothesis is that if you tried to scale the IAS model to researchers outside the top 1%, it would be less successful. I suspect the reason for this is that cross-pollination of ideas that you get from working with other researchers and graduate students, combined with the clarity of expression and understanding you're forced to acquire when teaching undergraduates is the reason why research universities have been so successful in advancing human knowledge. This isn't to say there is no place for other types of research institutions. However, I am skeptical of unsupported claims that there are "better" means of advancing human knowledge that are structured very differently from a research university.