I appreciate the last section's perspective: that giving to the causes best supported by evidence, though very helpful, does not guarantee 10x (social) returns since measurement is imperfect and some big returns are likely from things that aren't fully studied yet. Besides how that makes sense in its own right, it sort of pushes folks to recognize that that the many useful strategies the effective altruism community has come up with still aren't a silver bullet for aid effectiveness.
Recognizing there's no silver bullet seems especially relevant to me because there's a folk interpretation of "most social programs don't work" that's more like "we spend tens of billions of dollars trying to alleviate poverty; why isn't it gone now?". And _that_ question only seems to make sense because people misjudge the size of the challenge relative to the amount spent on it and the returns you can expect from that money.
The U.S.'s yearly global health budget, if spent entirely on sub-Saharan Africa, would be under $20/person, and you don't, say, raise average income $10k with $20, no matter how wisely or creatively you invest it. Demanding that kind of return from charity is like expecting every investment to perform like early money in Google, except, perhaps, that the impact of your demand is less on the investor and more on the global poor.
Separately, it seems like a shame to me that in fields where we do have strong signs of a high multiplier, like bednets or deworming, the opportunities haven't already been fully exploited by governments and billionaires. Some of it is that large spenders are making their choices through a process much different from someone like GiveWell's, but some is also that governments, for example, spend a lot less of their budgets on non-strategic global-health aid than most think (http://www.npr.org/sections/goatsandsoda/2015/02/10/38387558...