Thanks! That's interesting. I suspected there's a methodological flaw in assuming actual death rates from the sample of detected cases who died.

We need this voice of reason. Otherwise we're just panicking.