The problem with early estimates of COVID-19 mortality rate is that there were cases that went untested. Mortality appeared much higher than it actually is because only those who were symptomatic got tested. With increased testing we can now better estimate the upper bound for mortality, however, many states that already saw a peak in COVID cannot provide us the data we need. Only states like Florida and Arizona where COVID has not peaked can give us the comparisons we need.
Using the assumption that a COVID death will occur 2 weeks after testing positive, Arizona saw 131045 cases as of 7/15/20 and 3470 deaths as of 7/29/20. Arizona has an upper bound mortality rate of 2.6%. Florida saw 301387 cases as of 7/15/20 and 6332 deaths as of 7/29/20. Florida has an upper bound mortality rate of 2.1%.
An upper bound of 2.5% is still an overestimation because the calculation treats deaths/cases reported between February and May with the same weight as more recent deaths/cases.
The mortality rate still does not count asymptomatic cases, a number we wont know until there is wide spread antibody testing. Perhaps we should figure out just how many people have antibodies before we start sticking everyone with vaccines. Before we make huge policy decisions the American people deserve to know the true severity of COVID.
Sourse: Google 'coronavirus deaths' and use the data google uses to create their graphs.
it's less than half a percent ACHKSHUALLY
I wouldn't be surprised if the mortality rate is much lower, but I am making my comparisons in line with how other major medical publications are reporting mortality rate. https://coronavirus.jhu.edu/data/mortality
Also I am only looking at the US under the assumption that the numbers coming out of Florida and Arizona are the most accurate we are going to get right now.