Thanks for the input! You’re right, we misunderstood what that paper was saying. Looking into it some more, the most trustworthy number I can find is from the WHO, which says 11-16% diarrhea reduction from a protected well. They also say that “the health benefit is limited because these drinking-water sources may be microbially contaminated and because basic sanitation may not adequately protect the wider community from exposure to excreta.” This probably means that our mortality reduction estimate is also too high. 

If we use 13.5% (the middle of the range reported by the WHO) instead of 52%, we get 12 DALYs saved per 1000 people per year. With that number, we get a cost effectiveness of $25 per DALY. Which seems more reasonable than what we were getting before, and is roughly the same cost effectiveness as chlorination. Unless there are other errors in our calculation that we haven’t spotted yet. 

That's a really good point. I'm inclined to think that there's an asymmetry that tips the balance towards suffering, like maybe the fact that they only exist for fractions of a second is distressing, or maybe pleasure requires a more fine tuned structure. But it's hard to avoid anthropomorphizing the randomly generated brains, so my intuitions might not be correct. There's also the whole negative utilitarian vs total utilitarian debate. 

Great post! I think cosmic threats are definitely worth learning more about.

Only one thing bugged me. When you say the heat death of the universe is "the 3rd law of thermodynamics", it confused me because there's already a 3rd law of thermodynamics and, as far as I know, it has little to do with heat death.