People keep bringing up edge cases like this. If we look at strictly died/(recovered+died) the mortality is 6%. Obviously it's lower because there are recoverie

People keep bringing up edge cases like this. If we look at strictly died/(recovered+died) the mortality is 6%. Obviously it's lower because there are recoveries which are not recorded. I'd tend to believe that the statisticians putting out the 2% figure understand enough about disease dynamics for it to be accurate.