Unfortunately that figure contains selection bias - it is measured for those who die, but the figure for those who survive is missing.

That reduction is simply not achievable unless you can accurately predict who is going to die, which turns out to be very difficult to do.

The most significant health costs are for chronic conditions, and their ongoing consequent costs.

Your point is as non-sensical as saying we could save millions of dollars by killing 10% of the population at random.