1/ Crucial points about the NYS antibody study: Other studies have shown the first wave figures understate the true prevalence by 50-100%. Thus it is quite likely the real number of infected in NYC is in the range of 3 to 3.5 million, at a minimum...

2/ This implies an infection death rate of roughly 0.5 percent, or 1 in 200. However that figure is almost certainly overstated, for two reason. First, as @JohanGiesecke pointed out, the virus seems to spread most quickly through populations most at risk...

3/ So the fatality rates seem higher than they really are early on. Second, physicians increasingly agree that ventilators were overused early on in the epidemic and may have - unfortunately- increased the death rate. As physicians refine treatments and test drugs...

4/ The fatality rate should continue to drop. Overall, the fatality rate seems to be settling in the 0.25 to 0.4% range, with the average age of death roughly 80.

5/ Not everyone will be infected, but the US is a big country, so on a population-level basis those numbers may look big; worth remembering that in terms of life-years lost the #covid19 figures will be smaller than annual road accidents and far smaller than overdoses...