There are ~500 cities with population over one million, so I think we are going to be seeing a 500-year disaster hit a million-plus city about once a year.
Sampling a timeseries is not like sampling an ensemble. But nonetheless I agree we will probably see many more large-scale climate events than we have.