If new infections dropped to 0 today (100% effective quarantine), we would likely end up at 20-30kk or so dead.
How the *expletive deleted*ck do you get that math?
Right now, as of 0910EST March 29th, there have been 124,686 cases in the US. Of those, 2,191 have died, and 2,612 have recovered. So if new infections dropped to zero today, we'd have 119,883 cases left to have an outcome. You think this is suddenly going to jump to a 20% (taking the median of your range) mortality rate? That argument is retarded. Not going to happen.
That said it's hard to estimate where we are in the infection curve because our testing protocols are changing so rapidly. I suspect, with no real evidence, that we are closer to the top of actual infections then the "confirmed cases" curve would imply. If we keep it around 500,000 confirmed cases across the US, our hospital network will be able to mostly handle it, and we will be looking at 2-2.5% mortality. Which is still a lot of people. If we exceed health care capacity, we could see that mortality nudge into 4-6%. Which is a ton of people and on a par with the raw numbers that Jamis and Nick are tossing about.
I don't think we will see the almost 11% mortality Italy has because that hospital collapse required an acceleration rate of cases we don't have, and I think just won't get due to sheer landmass and population. NYC, LA, and the Puget Sound area may emulate it, but the country won't as a whole.