Its an s-curve with parameters that are changing by the hour and is sensitive to so many other variables. I’m definitely advocating as extreme a response as possible, but I see those numbers more as a pessimistic upper bound (necessary for people to take appropriate action).
In the last 5 days (data from CSSEGISandData) the total number of deaths (excluding china) has doubled from 20679 to 38798.
In the 5 days before that, the total number of deaths (excluding china) also doubled.
So worldwide (excluding china) the total number of deaths looks like this:
Unless things change really really quickly.