It seems that the model I used three years ago was correct, but the parameters were not. This graph shows the cumulative number of Omicron infections in Beijing, which matches my model, reaching a maximum value in 50 days. Almost all of Beijing's 22 million people have been infected.
