Lots of reasons I don't think you can extrapolate meaningful death rates from this yet. One big one is that time from infection to death covers a very wide ran

Lots of reasons I don't think you can extrapolate meaningful death rates from this yet. One big one is that time from infection to death covers a very wide range: in Wuhan time from hospitalization to death had both a mean and a standard deviation of about 2 weeks, and time from infection to hospitalization isn't well known yet, but is likely at least a week. Add to that the fact that some cases are diagnosed post-mortem and you might not know about all of the fatalities for early April cases until well into May.