My perspective is, who cares how many people have the virus. The important question is how many get seriously ill.
But the seriously ill-to-infected ratio is also important from a general policy-making perspective, for making good decisions on how the general population should behave.
Hmmm, possibly a lagging indicator type of situation? The deaths are in patients diagnosed a week or more ago, where the case counts are current. If by mortality rate you mean number of deaths divided by current case count, then you are using the wrong denominator? It should be the number of cases reported at the time the patients who were diagnosed, not the current number?