This research comes from data collected in the earlier days of the outbreak in China. It's 3 months later and things are completely different now with better understanding and social distancing, so the R naught should be lower. Again, I don't understand CDC's decision to publish this on their website. The media is going to take this and run more fear mongering journalism with it.
I did do believe the current estimate of 2.2 is on the low end. This thing is just as infectious as the flu (R naught of 2), but with more than double the incubation period, and you can still spread it during incubation and asymptomatically. This virus will be very difficult to contain, so all we can do is flatten the curve, approve more treatments and get herd immunity until we can get a vaccine. Yes, there is a possibility of repeat infection, but we need more solid data on the percentage. Virus mechanism wise herd immunity should work, like SARS and similar to some communities like Copenhagen during the Spanish flu.
You know what else would help cut down the R naught CDC? More people wearing masks to prevent the spread.
To derive realistic values of R0, we used previous estimates of serial intervals for COVID-19. The serial interval is estimated to be ≈7–8 days based on data collected early in the outbreak in Wuhan (1). More recent data collected in Shenzhen Province, China, suggests that the serial interval is dependent on the time to hospital isolation (Q. Bi et al., unpub. data, https://doi.org/10.1101/2020.03.03.20028423External Link).
They used a more recent data.