There are any number of articles on this subject in “Lancet” and the “New England Journal of Medicine”. They are all open access. I use the source below from Johns Hopkins. It breaks the data out by epicenter, which is similar to a cohort time approach. This is where the 1 to 2% figures come from.

https://gisanddata.maps.arcgis.com/apps/opsdashboard/index.html#/bda7594740fd40299423467b48e9ecf6

The one thing that is important to keep in mind: the case definition only counts symptomatic people. So the mortality rate is mortality among people who present to a hospital with symptoms for testing. We simply do not know the denominator as we do not know the rate of asymptomatic transmissions.

The incidence and mortality data from Italy, Japan and South Korea are by now big enough and reliable enough to serve as a reasonable sample-size to estimate mortality rates under best clinical care. I trust their reporting system.

How do your calculations look when applied to these data sets.

Dirk