Because the IHME model is based on taking China's data at 100% face value and applying it to the US. It over-predicts in the short term and under-predicts in the medium to long term. They keep revising the overall number down to line up with an epidemic peak in 3-4 days in the US, which is ridiculous.
The reason why the epidemic peak is estimated so soon is that New York has the bulk of the cases and they began bringing new cases under control before most of the rest of the nation.
Does anyone else think NY called their peak prematurely? It seems like there's some periodicity in the data and there's always a slight two day drop every 7 days. This makes sense since many people are still working mon-fri in the city.
The actual assumption is that with current social distancing, any outbreak will die down on its own on approximable dynamics. There is good evidence that this is so.
All right, that seems much more reasonable to me. Thought you said the peak for the USA in general was imminent, and that's not at all obvious given that the social distancing policies going forward have yet to be decided.
It might be the case, but that assumes that effective measures to control the epidemic will keep being applied everywhere.
