How about giving the institutions very top-level goals, on say 20 or 50-year basis, and then just give them a bag of money? Research is a high stakes, high rew

How about giving the institutions very top-level goals, on say 20 or 50-year basis, and then just give them a bag of money? Research is a high stakes, high rewards type of game. You need to pursue many potential ways forward, and most likely only one will win. Rewarding short term wins makes academia no different from industry, and prevents finding anything but the local maxima.