Keeping in mind these are fat-tailed distributions. So an R naught of 2 is a estimate of a true R naught, and because using sample averages to estimate means (true R naught is a mean) is unstable (the fatter the tail, the less moments of the distribution there are), once the reported number starts getting upwards of 2, the less confidence you can have that that reported number represents the true R naught. You have know idea if the superspreaders are infecting 50 or 500 people.
At the beginning of the pandemic, we were seeing R naughts estimates at 2.5. That's shockingly high. People were right to take extreme members in the face of that kind of uncertainty.
These people enforcing their values with 20-20 hindsight when the decisions were made given limited data at the time are causing harm to our collective ability to make decisions in emergencies.