coronavirus math

Interesting analysis (in your referenced paper.)  But a question that I didn't see answered in the paper: 

The authors define population sizes {N = S(t)+I(t)+R(t)} for each of their models, but those numbers don't agree in any coherent sense with the country's population.

(If you let me "turn enough knobs" I'll match any trend.  😉

It is, however, a very interesting discussion:

Most likely, the reason for the
outbreak was that at the beginning of the pandemic, the
United States decided to follow the path of rapid
immunization, i.e. obtaining collective immunity due to a
greater number of patients. The decision not to impose total
quarantine was due to concerns about the economy.

Those of us sitting in the middle of this debacle would put a different spin on it, your authors were kind enough to assign at least a conscious choice to our federal government.

It is interesting that the percentages of your paper's populations are so low compared to the actual populations.  There are at least four studies that demonstrate that a significant number of people are infected and can transmit the virus but display few or no symptoms, are "unsymptomatic."  Two of these studies claim that 50% of the people shown to be infected have no symptoms.  With that in mind I built (yet another) model to address that.  All of my models to date result in much higher rates of infection until I set the symptomatic percentage of the population to a very low (1.67%) value.  A rough working copy of this latest model (in Prime 4) is attached.  Data from the USA is embedded and compared.  The effects of "social distancing" are included.  And the possible repercussions of relaxing that lockdown can be evaluated.

I solicit and appreciate comments and criticisms.