Oh, I get it now. You know more than the experts who do this as a career.

You say “ As it turns out, actual hospital bed utilization is way below IHME predictions. The closest model might be Florida where actual beds needed / model forecast is 47%. In New York it’s about 36%.”
Where did you get this gem of a factoid? (no citation). Don’t believe it. And the sick and dying health care workers in NYC might disagree with your skepticism.

As for models, you say “ They haven’t updated any of their models.”
Wrong.
We’re getting real time data now from hospitals. Data is used to update assumptions made when a model is designed. Build model, make best assumptions with data you have, as new data tolls in you update model and it gets more accurate. That’s how modeling works.
The article you linked actually states:
“IHME will update the analysis routinely as a source of “real-time” information for hospital system administrators and local, state, and federal health policymakers. The analysis is based on an extensive range of information and data sources, including:

-Local governments, national governments, and the World Health Organization
Government declarations on implementation of social distancing policies
-Data from the American Hospital Association on licensed hospital bed and ICU capacity and average annual utilization of these services by state
-Data on COVID-19-related demand for ICU services in the United States
-Age-specific death rate data from China, Italy, South Korea, and the US”

And the article linked within that parent article further states:
“The model was last updated at 7 a.m. Pacific, April 2, 2020 and will be updated again on Saturday, April 4.“

I’m not sure what your angle is in all of this. But keep it coming. Thanks for linking to that article.