On its face, this is probably true. But when the process is showing exponential growth, and quality data (i.e., deaths) is lagging by weeks, it's not unreasona

On its face, this is probably true. But when the process is showing exponential growth, and quality data (i.e., deaths) is lagging by weeks, it's not unreasonable to be very conservative in one's risk estimate. Once the exponential phase has clearly passed, I'm guessing we'll start treating this more like traffic fatalities. As in, Yeah, tens of thousands die each year, and we could easily cut this in half, but why bother?