A very thoughtful set of notes. But as you yourself point out, it’s more of a debunking of the sketchy nature of the current thinking than a prediction that pretends to accuracy. I applaud that too — too few people are up front about the limits of this kind of back-of-the-envelope calculation. I think there’s a flaw in the logic though. If it’s 6% of everyone needs hospitalization, god help us, but most of the serious illness appears to originate among a small slice of the population. Not all, but most. Say, for instance that almost all of the serious complications (including death), i.e., the red portion of your graph, are among people who are over fifty and either have some pre-existing lung problem, COPD, etc., or are immune-compromised. If 25% of the population is over fifty, and of them, .15 of them have a relevant preexisting condition, etc, then the area under the red curve is drawn from 0.0375 of the population, not 100%. Now say 75% of all of us get it in the next year and for simplicity, that there is no difference between the probability of the young and the old getting it. The portion of the public from which the severely ill patients are drawn would be about 4% of the size the graph assumes. Now say 50% of the people in that group get it bad enough to be under the red curve (as opposed to 6% of the whole population as shown.) That’s about 14 people in a thousand who would be hospitalized over the year. If 1/4 the cases happened during the worst months that’s 3.5 people per thousand during that surge. This would exceed capacity if there are a total of 2.8 beds per 1000. but that’s the point of “flattening the curve.” Say our medical system can afford adding 1 person per thousand to the current load of people sick enough to be in the hospital by sending the not-too-sick home and adding some beds in the halls, etc. Even under the fairly dire assumption that half the people in the high-risk group need hospitalization over the year, you only have to slow down the epidemic by 3.15 and Bob’s your uncle. I think the scenario as described is pretty plausible but even if it’s too sunny, it does not sound like you have to “flatten the curve” all that much to make a big difference in the gap between the number of sick and the number of beds.