We must be under-counting mild cases of flu, too, right?

Surely lots of folks with milder cases are never even diagnosed?

So the denominator for the death rate is either a guess or an under-count.