Agree with you it is a kick starter and with a few tweaks he could have had a stellar op-ed, but where there is a bias I will give him 100% credit on consistency.
Back in March I had to make a decision about whether to go fly and see my parents 50th which would be two weeks from now. To do the swag, I looked at the mortality rates of the flu and saw the transmission rate was 4x as high as the flu (22–25K a year). This led me to believe that we would see around 100K deaths this years if we followed a single wave and mitigation was successful and held until April 30th.
So I used the 21 day theory (using the fact that the virus survived for 19 days on the surface) and combined it with the observation of wuhan where it looked like it took 4 cycles to reach the peak. and determined that on March 26th …. I couldn’t fly until August. No trip for me ….
Where my modeling (remember this was march) went bad was the fact that social mitigation was not done immediately and it was hap hazard and the country lacked the effective medical resources to implement it (we literally outsourced it all). It presumed a single vector Asia > USA and not the double point (Europe > USA). So in reality I needed to double it (so my view should have been around 200K on the estimate) for this year …. But It was good enough to prevent me from flying to what is the start of a second wave due to a combined re-opening and social protests (no complaints from me on that aspect) …..
To answer your question though ….
To break this down, I would need to look and see what the total death expectation would be (initial proof would need to validate the 0.86% death rate (3million or so) and the herd immunity requirement (70% initial estimate) and then start to build models around how fast it spread, when a vaccine could be implemented, what is the probability of herd immunity being gained by exposure, the probability of strain mutation, the effects of population demographics on the model.
I would then have to work in other aspects …. e.g. I can look at a test case of locality in my area, and then compare it against Dallas were they had protests for 10 days straight to see what models can be built on transmission.
E.g. I can look at my county (Grayson) [north of Dallas] and see that two people brought in and caused a long running limit of 34 cases. Then one person caught it at a Tyson plant and spread it to 300 out of 1700 and from there we had 4 deaths. So we can see in a rural sparse area with a high risk section …. 2 lead to 34 with a low test rate and then we had a large spike 300 leads to what … (we shall see).
It was a pleasure discussing things with you in a civil manner. I know I didn’t answer your question, nor could I … but it is something I could work out as long as my job gives me the time to do so :)
Sample link for Texas and it’s spread
https://txdshs.maps.arcgis.com/apps/opsdashboard/index.html#/ed483ecd702b4298ab01e8b9cafc8b83
Great article on this virus.
https://towardsdatascience.com/landscape-of-coronavirus-research-across-three-decades-9cf338968e18
