I'm a data guy...I can't help myself.
We will probably pass 60,000 deaths in the next 24-48 hours. That shows a death rate of about 6%. However, we know that there is a severe test shortage, and initial results from Iceland, USC, NY, and South Korea show a much lower death rate than that. This probably means there have been over 3M people infected.
I'm a data guy too... While this is a morbid subject, from a data science perspective it has been fascinating.
Obviously the death rate based on confirmed cases is nowhere near the true mortality. We've known from the start (and people haven't hidden the fact) that due to mild/asymptomatic cases, none of the "ultra scary" numbers the media threw out made sense.
Right now the few US antibody tests show a pretty wide array of results, and what that means for the true mortality rate is interesting.
CA shows somewhere in the realm of 40x unconfirmed to confirmed case ratio. If you applied that to the 6% death rate, you'd end up somewhere in the 0.15% true mortality. If you compare it to California's death rate (3.9%), it would bring you under 0.1% true mortality. If that rate is accurate, then we should let 'er rip and open the country up.
NY showed differently. Antibody testing only suggested a 10x ratio of unconfirmed to confirmed cases. That extrapolated out to somewhere in the range of 0.5-0.78% true mortality (depending on whether you include the "probable" cases from NYC). If that rate is accurate, then given how transmissible this appears to be, a let 'er rip response would be several hundred thousand more deaths, possibly in the 1M range depending on the total number of Americans who ended up contracting it.
South Korea was WAY out in front of this thing and has tested, and tested, and tested. With over 600K tests administered, they've only found 10,738 cases, and had 243 deaths. That's a 2.3% mortality rate, which is a lot higher than most predict is the true mortality rate. So it's expected that they have a much lower ratio of unconfirmed to confirmed cases based on their test rates, but without antibody testing we won't know who was missed. But if the true mortality rate was merely half of that, it would be devastating.
I wish we had better data all around, because we're making HUGE decisions about our economies and there are boatloads of unknowns.