I was hoping they would introduce concepts like Kalman filtering, that address the problem of unknown variables and other uncertainties which affect your results. The models that I’ve seen all seem to pretend that they’ve accounted for everything. Everything that’s important, anyway.
What I think the authors were trying to show is that there are ways of easing back the restrictions so that the R0 rate remains low. Also, that binary restrictions/no restrictions are not a good idea.