This rule lines up with age group and preexisting condition groups surprisingly well. It underestimates the death rate of older people, but is within a factor of two. It drastically over-estimates the death rate for children.
Corollary 1: If we do nothing, then hospitals and morgues will have to absorb a year worth of cases in about 2 months. Obviously, that’s a big problem.
Corollary 2: COVID will reduce life expectancies by much less than 6 months (actually, less than one month).
I suspect we could increase life expectancies in the US by much more than 6 months if we spent $3T on something other than the shutdown.
I think Corollary 1 assumes 100% infection at T_0, which is clearly not true. It's debatable what the correct curve would look like, but even in the worst do-nothing case I don't think it could possibly be this steep.
I don't understand the logic for Corollary 2. Even if we assume Corollary 1 (which I'd argue we shouldn't) why would uniformly doubling the average death rate reduce life expectancy by less than 1 month? I can't say for sure that it wouldn't, but I don't think the math is obvious. Could you flesh this out?
Your final sentence seems possible, depending on what "much more" means, but I'd like to see more proof. Adding $3T would be equivalent to doubling current US health care spending.