I'll probably rewatch the questions section of the webinar for anything important to us. One thing I picked up was that someone was expressing concern about how precise the estimates might be (which fits my narrative!) and wanted some measure of uncertainty. So I think providing distributions over new infections will be good to do. This will need standard errors on the the RRs, maybe draws from the Naomi model or just treat it like a Gaussian and use the standard error, ..., then Monte Carlo it all together to get uncertainty out the other side. Uncertainty on those "infections averted" graphs I posted earlier would also be good.