In fact, though, local behavior _does_ make a difference in the evolution of epidemics. For example, just look at the variation in epidemic characteristics across countries for COVID-19.
Perhaps one way to get at this would be to analyze a population into local zones, where each zone has its own \\(S, I, R\\) populations. Each zone could have its own transition coefficients, depending on local conditions.
Then there would also be transitions representing the flow of individuals due to travel between zones.
Not sure whether this has been done, or whether it would be fruitful, but it's a concept.