Agreed that the naive method will lead to underestimation of the CFR. Though I'm not sure if it's really as high as estimated here. Looking at the numbers rep

Agreed that the naive method will lead to underestimation of the CFR. Though I'm not sure if it's really as high as estimated here. Looking at the numbers reported from provinces in China other than Hubei (from https://en.wikipedia.org/wiki/Timeline_of_the_2019%E2%80%9320_Wuhan_coronavirus_outbreak https://en.wikipedia.org/wiki/Timeline_of_the_2019%E2%80%932...), as of 5 Feb 2020 there have been 14 reported deaths but 520 reported recoveries. This puts the CFR at 14/534 = 2.6%. I can only guess why the CFR in Hubei seems to be so much larger. My guess is that the medical system in Hubei is simply overwhelmed by the number of patients, and that only severe cases get treated in hospital and diagnosed. Patients with mild symptoms don't get diagnosed and the majority of them recover. Still, caution is advised. Everybody take care and be safe!