[scarmig]

~0.5% is a very reasonable estimate of the initial CFR when an epidemic starts. It's bad, around 5x as much as the flu, but probably not worth a national China-style quarantine.

But it has a key context: that's the CFR for when all patients are treated. In most of China, there were never enough infections to trigger a healthcare system collapse: that's why in all but one province, the CFR was indeed ~0.4%.

In one province, Hubei, the medical system did collapse.

Its CFR jumped to 3-4%. That is worth going to extreme measures to prevent. In the USA, that would amount to ~10M dead.

That is not a seasonal flu.

I think a superior approach for most of the world is one along the lines of Singapore, Taiwan, and South Korea: extensive testing and contact tracing and limited but real quarantine measures. But, yes, everything should be on the table to prevent the nightmare healthcare system collapse scenario.