@Alon it’s a hard thing in general that what would be “high impact” may well correlate with what would require subtle forms of institutional capacity. both reputational defensiveness and a kind of scientism that prefers replicability according to public recipes would tilt away from that.