(1) Yes, the evolutionary pressures are possible. There's also the related notion of treatments getting better over time, etc. Basically, and you alluded to th

(1) Yes, the evolutionary pressures are possible. There's also the related notion of treatments getting better over time, etc. Basically, and you alluded to this, my position is those effects may occur to some extent, but will be massively dwarfed by the downsides of trying to slow down the timeline. (2) As far as the CDC, I hate to be that guy but I don't trust their numbers very much. I'm just a bit suspicious of them upwardly revising their number at this point in time, especially given the statements coming out of the CDC head which seem to lack any type of nuance. In particular, they are modelling asymptomatic spread but not pre-symptomatic? Weird. But yes, I generally give a range of .1-.7%, but settled on .30% as the best number for my purposes. It's more-or-less pulled out of my ass, although it is almost exactly the median in the big spreadsheet of IFR estimates which I now cannot find the link to (grr) Anyway, the important thing is for my "argument", it actually doesn't matter to me if the IFR is .3% or .9%. The costs of lockdown just end up being too risky. So, feel free to ignore my mistrust of the CDC. Note that Ferguson's classic (albeit incredibly myopic) paper models an IFR of .9% and 82% of the US population being infected, with an end result of 2.2 million. With these T-cell findings and other findings, we might see like 1/4 as many total infections, so even with the same IFR that gives us a shitty upper bound of 600k.