Deaths and severe cases doubling period roughly 3 days during the early, exponential phase.
I think this gives us a good, rather unbiased proxy of actual epidemic doubling time.
After all, when assessing doubling periods it does not matter if there is systemtic under-counting of cases (i.e. if cases are off by a certain fixed factor) the doubling time will be the same.
Also consistent with latest research.
Former research (Leung et. al) used to rely on case doubling estimates of 7 days.
Many people have difficulty understanding nature of exponential growth.
Case doubling of 3 days, means: epidemic is one thousnad times larger after just 4 weeks and one million times larger after 8 weeks.