Large-scale antibody test suggests Covid-19 cases underreported by factor of 50

Math hints to this anyway. Isolated cases like Diamond Princess do show around a one percent death rate. And death is a trailing indicator by about 2 weeks. In 2 weeks the number of new daily cases goes up by up to 20x in hot areas. So to get an accurate estimate of cases you'd multiply the number of deaths by 2000. Currently the number of official cases is only 20x the death count. So, the result is not surprising.

This is also somewhat fantastic, as it means 1/6 of the US population has already had the disease, and maybe herd immunity is not a distant dream. (Though if any of this is true then that also means the virus is more contagious than previously thought, and the herd bar is even higher than the 75% experts have been mostly saying).