That's a somewhat different thing, Steve. It is a bit like the difference between sin(x) and sin^-1(x). We would not trying to work out what the herd immunity threshold (HIT) is from the R, as we have to at the moment, where

HIT= 1 - 1/R₀

That is really difficult because R₀ has to be estimated from all the local R values which vary across the country and are affected by policies like tiers and lockdowns. That reciprocal also means quite a small change in the R value can have a large change in the HIT, which is why we cannot be very certain what the range of the percentage for herd immunity is.

Rather, we are working from the opposite end: given we have a certain level of immunity (via infections and vaccines), what effect is that having on R?

It is admittedly a little over 40 years since I worked on this sort of stuff, but as far as I can tell the fundamental modelling hasn't changed that much.   In those days we had to model essentially a population and now the modelling is more based on large numbers of interacting populations, so that adds a lot of variation and sophistication I haven't worked with at all, but the heart of modelling does seem to be similar.