I did the calculations on another thread. The percentage of deaths vs. total cases is way too close for the latest days (1/28 to 2/2) to be natural. Most of them were 2.16% to 2.17%, with one something like 2.12%.
With any natural set of data there will be more randomness. It would be interesting to see how the “official” data compares to various models.
You don’t shut numerous cities for 200 dead. (The number of dead when they started doing the quarantines.)
You're right - the numbers don't 'add up'...