[termain]

That's generally what you should do with a scientific model: state your assumptions and methodology clearly.

And most if not all scientific and engineering models spread properties across a group of objects and use averages. R_0 is an average. And it's the average behavior that matters here

In a well tested population, you might be able to focus interventions on infected people and their contacts. In the US population, the people most likely to spread the disease are those who interact most with other people. That means getting those people to reduce their interaction rate.

It's a semi-empirical model, as are many if not most models used to make decisions.