> This is an infinite benefit. So to model that without getting nonsense out you have to use a discounting model where you arbitrarily consider infinite future value to not be worth more than, say, 10 billion dollars, and there's your answer. But that's a paper exercise, why bother? You could choose values that say a trillion dollars or $9.99.

It's really not that arbitrary. There are many ways of valuing this, sure. Here's a fairly straightforward one:

Say it costs you A per year to maintain current levels, and the yearly cost of capital to eradicate is X amount of capital * y% yearly cost per year (e.g. if you borrow X at y% interest, or if you have X but sacrifice y% returns by spending the money instead of investing it), and the lowest viable cost to save at least as many lives as the infection currently claims is B, then it's a question of whether A+B or X*y% is lower. Whichever is lower will allow you to divert more funds to other targets, and so save more lives (or do something else). Of course there are risks here, in that any of the variables can change, so if they're close it may be worth it to eradicate simply to eliminate the risks that the alternatives get more expensive going forwards.

But this is not a very hard thing to model. The reason we don't do that more often is that often there are many other factors involved, such as politics and emotions and who fights hardest to get funded.