I definitely don't think we have a definitive answer yet. I think one day you might be able to get a good definitiveish answer. For instance, you can assume th

I definitely don't think we have a definitive answer yet. I think one day you might be able to get a good definitiveish answer. For instance, you can assume that everyone gets covid in the "no measures" case, and extrapolate death rates using serosurveillance to get the portion of the population infected in reality. Once Covid is "over" you can see the number of people who died in reality. There will always be all sorts of errors in such an estimate. From bad data, to things like death rate depending on your assumptions about variants, vaccines, and the make up of the population. But once we get the "how we did in the end numbers" it's plausible that we can make some sort of comparison to "how we could have done".