I see now where you get the 9%, but I still think it’s an overestimate.
First, that resolved-cases calculation combines the more recent low death rate with the April horrors in the NYC area. That death rate in NYC was partly the hospitals being slammed, which will happen again - is starting to happen again - in other places as they let the infection spread widely.
But it was partly the newness of the infection. By now, we have better information on how to use ventilators, how to use pronation, etc.
Even if FL, TX, AZ, etc. have failed to learn from NY on controlling the infection, they’re likely to learn from us on treatment.
Lastly, there’s the issue of undiscovered cases.
The numerator of the closed-cases calculation is probably missing some Covid deaths.
But the denominator much more likely includes lots of infections that are uncounted because the person has no symptoms, or are reluctant to get tested, or they find that testing is not available to them.
So there are lots of people who got infected - and maybe sick and recovered - who aren’t in the denominator of the closed-cases number.