One possibility is that instead of modelling 'how contagious is the disease', which can have assumptions that the population is homogeneous in susceptibility, we should be modelling the susceptibility. It's instructive to look at 2020 figures for serious illness and death in South East Asia -- Vietnam, Japan, South Korea, China, Laos, Taiwan. Hardly any cases, and hardly any deaths. The countries that were seriously into masking and lockdowns claimed that this was the reason for their success, but other south-east Asian countries who didn't do much -- including Japan for most of 2020 -- also were also spared.

This looks like 'south-east Asians were resistant to covid in 2020' to me. So you then wonder -- is it their genes, or their environment, or prior contact with some other disease that gave them immunity? Somebody did a study in Thailand -- and unfortunately I cannot find it now -- which seemed to indicate that new immigrants (in the last 10 years) were much more likely to catch covid than those who immigrated from longer ago. They were looking for socio-and-economic factors related to poverty and income, but I said to myself 'looks like everybody got mildly sick with something more than 10 years ago, and it's protecting them now'.

Now we get to 2021. Along comes the delta variant. Cases spike in south-east Asia. Whatever was giving them resistance to catching earlier variants of covid does not work for delta, it seems.

I now wonder how much of the theory I learned in school 'over time, diseases mutate and become less lethal' is driven not by changes in the disease but in the susceptibility.