You've got to multiply that with the probabilities in various somewhat complicated ways, but a simple approximation is to multiply the probability of death by a

You've got to multiply that with the probabilities in various somewhat complicated ways, but a simple approximation is to multiply the probability of death by a certain kind of cancer by the expected loss of dying of that cancer at the expected age to die from that cancer. Yes, it's subjective. No, that's not fixable, and in my opinion it's not a problem.