Zuck seems to think that just because he's brilliant with computers (and making money with computers), he's brilliant at other things.
It takes some finesse he doesn't have.
There's two real requirements to fixing arbitrary massive problems in the world: be a polymath, and understand where and how to leverage effort. Being a polymath means you actually have to approach new and interesting problems by learning about them; and that learning will always be incomplete, so you have to take what you know and lean it against people who can pick it apart.
Zuckerberg wants to solve disease. I want to solve a great many of the United States's economics problems in one move--poverty, economic stability, technological growth, and the broken welfare system. These are both problems we're not experts in; how do the approaches compare?
Zuckerberg's answer is, to summarize, "I have all this money, so I'm going to pay all these smart people to fix disease forever. Incoming cancer cure!" Buy into your own research center and have them figure it out.
My answer involved a lot of ruminating on the way economies function, on monetary policy, on the source of wages (it's revenue, which is spending, which is income), and so forth. I inspected the government's spending, their taxes, and the income distribution in America. I identified problems in the welfare system. I identified risks--business risks in the end point, transitional risks getting from here to there--and created mitigations and contingencies to handle them. In the end, I came up with a Universal Social Security plan; a transitional phase to get from here to there; and explanations on how to lower landlord risk so that the stable income the USS supplies is enough for landlords to profit by renting, largely by reducing the risk of evictions and empty units, thus reducing the cost-of-risk and associated high rent costs required to maintain profitability.
If I had Zuckerberg's resources, I could take this plan to a campaign. I could get the attention of economics researchers, congressional staff, and eventually Congress. Instead of standing up and saying, "Hey, let's solve poverty! We're all rich here, it can't be hard. Maybe we could give out scholarships so all these people can get job skills and start working!", I've developed a plan that accounts for the basic facts of economics and the current economic situation, identifies a method of handling the problem, and avoids and controls the negative consequences of those actions.
If Zuckerberg wants to cure a disease with his freaking billions, he needs to find a disease into which he can supply some profound insight. That's going to take months or years of pouring over the problem himself, not small billions of dollars sunk into yet another research shop. This is the kind of thing you see Hopkins students pulling out of their asses; if we knew how to pick those specific individuals out of population just as their brilliant ideas were forming, we'd put them all in a room together and solve all disease in like $4 million. That happens to be completely impossible.