That's not what that is saying.
It is saying they estimate 56k beds will be in use on the peak day.
I agree that you cannot take each day's resource use, add them up, and get total people hospitalized over the course of wave 1. Because people use a bed for more than 1 day.
Without knowing the average time a person is hospitalized, you cannot convert the distribution curve of resource use from IHME to the cumulative hospitalizations curve for comparison to the 490k total (cumulative) hospitalizations (or hospital beds) used.
However, we do have data on average time of hospitalization for flu. So, I used that and information about hospitalization rate during peak flu season to estimate the peak resource (i.e., hospital bed) use for flu.