This is an absolutely brilliant way of communicating the impact of assumptions on the outcomes of the models. The sad truth about our COVID data today is that the error bars are so wide on most key assumptions that best case, everything is ok, worst case, it's Armageddon.