«it's highly likely that there are more unreported cases than unreported deaths, which means that the actual mortality rate is lower.» I don't think so, becaus

«it's highly likely that there are more unreported cases than unreported deaths, which means that the actual mortality rate is lower.» I don't think so, because deaths lag infections. When the outbreak is rapidly growing, if it takes about 5 days for the virus to kill, then even if you knew all the stats perfectly, you actually need to look at the ratio of "deaths as of day D" over "infections as of D-5" to calculate the mortality rate. If you do this math right now (291 infections as of Jan 20, 56 deaths as of Jan 25) you find the mortality rate is 19%. Assuming infections are underreported by a factor of 5, the mortality rate is still high at 19/5 = 3.8%.