It would undoubtedly be interesting to see life expectancy by income. I found this and this (and google gives you plenty more). Briefly, there is a pretty clear correlation between income and life expectancy in the US, and, unsurprisingly, the gap between the life expectancy of the very rich and the very poor is substantial (more than a decade).
I’m not sure it’s fair to blame (rising) income inequality (and life expectancy inequality) in the US on the taxation structure. Probably it plays some role, but I think a much larger role is played by the social structure: pro-corporate and anti-union sentiment, political corruption in the form of crony capitalism, lack of universal health care, etc. But, obviously, that is a much bigger debate.
For instance, if we look at Canada, we see a lower gap between the highest and lowest quintiles: roughly 5 years. Canada has (I think) lower income inequality, but also universal health care. I don’t think its obvious how much each contributes to the more level life expectancy curves.
Agreed that the devil is very much in the details.