First of all, you can't calculate a fatality rate without a base rate of cases -- without widespread testing, the base rate is likely under-estimated.

Second, the vast majority of deaths (I believe 75% of them?) have occurred in China. China has terrible air pollution and a much higher proportion of the population that smokes. COVID-19 is most dangerous in those with preexisting respiratory issues. There is a sampling bias happening there, I think.

So, it might be correct to say that if you filled a stadium with 50K people randomly selected, you would expect 10 deaths from flu. I don't think it is accurate to say you would expect between 500 and 1500 deaths from COVID-19. That death rate would likely be accurate if you had a disproportionate percentage of your 50K people who had preexisting respiratory issues.

I think the containment efforts are important, but there is a freak out level that is just strange.