A crude mathematical model from Swedish authorities on the Stockholm outbreak used two parameters: 1: the fraction of undetected cases (assumed to be mild or asymptomatic) and 2: the relative infectiousness of that group compared to the “detected” group.

The larger the undetected group is, the lower their relative infectiousness has to be in order to fit the observations. The best fit I believe was 1/25 detected and 11% infectiousness of the undetected group.

For the particular conditions in Stockholm, such as testing prevalence etc

Yes absolutely. The actual parameters would be different everywhere but they seem to indicate that the symptomatic group is more infectious (which I guess is the base hypothesis for a droplet transmitted disease).

Thanks - that's the kind of thing I was looking for, but this line stuck out to me: "However, the evidence of the relationship is limited by the poor quality of many of the studies, the retrospective nature of the studies, small sample sizes and the potential problem with selection bias." The book I mention gives me enough reason to doubt that what we know about Covid-19 at this point is anything like the whole story.

I actually do remember reading that this was true specifically in the case of COVID — that more exposure so far seems to correlate with a worse infection.

Sadly, I have no idea where I read this. But... I know I did! Recently! Maybe NYT?

This is the article you meant: https://www.nytimes.com/2020/04/01/opinion/coronavirus-viral...

Yeah, I saw it too. The hypothesis, as I recall, is that the more virions inhaled, the more likely some of them will get deep in the lungs where they can do the most damage.

I read something like that too. I came away thinking that ingestion might be a better way to get it than inhalation. I think you really need to keep it out of the lungs and nervous system. But that's all my impression from who knows where.