I will argue, even if the true intervention has a varying effect, it’s sorta-mostly OK to just take a hazard ratio from a paper and convert it to a change in life expectancy using this curve

This assume the interventions are independent.

This math salad is exactly why weak hazard ratios should not be used for personal health advice, but to form a falsifiable hypothesis and do a actual interventional test.

You cannot inform on cause and effect from weak observational "signals" and no amount of mathematical gymnastics changes that. You actually have to perform experiments