You just proved my point. If they hadn't used the most ridiculous predictions the first time around, maybe more people would take their predictions seriously.
Models aren't meant to predict the future. Models are meant to predict what will happen if things follow along in a certain path. There are models for many different potential paths. If a path changes, the model would not predict the same results, eg. if a cure was found a week after the model predicted 400k deaths, the model would be outdated pretty quickly. When the model predicted 400k deaths, it was assuming that policy and culture will pretty much continue on the same course, and unfortunately our actions as nation helped make that a reality.
While the epidemiologists can show models that predict 0 deaths and not scare anyone, we should want the models that show what happens when nothing changes. Your criticism of the models has been debunked over and over, this time very much so.