One possible issue here is that just because a cause isn't "provable", it still may be worth doing if it has a good chance of having a positive effect.

For an example, say you're running a foundation and have to decide how much money to give to mosquito nets now, and how much to invest in research to develop a malaria vaccine. Of course it's not "proven" that trying to develop a malaria vaccine is going to help people, and it can't be proven until it succeeds, but it is quite possible that it might. And if it does, it is likely to have a significantly larger effect then nets.

In other words; provability is certainty one issue to consider, but it's not the only one.