You're assuming that research is fungible. That, say, if someone is really good at developing novel, efficient data structures, they'll be just as good at finding ways to help homeless people. Why do you think that's the case? Nothing in the real world suggests it works like that.