I think one of the issues with these analyses of the fatality rate is that it is variable and clearly increases when medical systems are overloaded. This creates an issue with analyses like these which are "technically" correct but don't take into account the fatality rate would be higher if we were not successful in "flattening the curve."

What it suggests is that to monitor your own safety you should keep track of number of cases in your area as well as medical resource usage.