Two significant notes, however:
* China reports that its outbreak has largely resolved. The vast majority of the 1,000 cases in Iceland are still active, so we don't know how many deaths Iceland will eventually have. South Korea's current fatality rate still seems to be around 1%, with many cases yet to resolve.
* China seems to include only symptomatic cases in its total. That should bias its fatality rate upwards, since the denominator does not include any asymptomatic people who resolved without developing a full case of COVID-19. (Note that this will also be true for jurisdictions that require symptoms for testing. I believe that SK tested many asymptomatic people, but the US for example is not.)
Note that the 1%CFR estimates are based on full access to medical resources. When medical systems are overwhelmed as in Italy, we expect that number to increase from that baseline. Jurisdictions that report a markedly lower fatality rate are likely to have had systematic problems in identifying or reporting on case totals.
2820 positive cases.
Positive rate: 0.8%
Iceland: 3,787 tests by March 18,
330 positive cases.
Positive rate: 8.7%
Clearly Guangdong did more testing with respect to its epidemic than Iceland.
Let's look at New York State:
172,360 tests by March 29,
59513 positive cases
Positive rate: 34.5%
Hubei population is 60M. So they'd have to do 2.7 Million tests to have the same degree of testing at Iceland.
Even with an excessive lockdown, it seems that just on this issue alone, China's numbers look really quite wrong.
That there was a breakout in other parts of the country and almost no deaths look very suspicious.
Combined with the fact that all public Chinese information is part of an 'orchestration of public opinion' (i.e. propaganda effort) and is not based in any reality, so of course, there's a credibility problem to begin with.
So you are saying that countries and regions that are under-testing relative to Iceland all have numbers that are wrong?
"That there was a breakout in other parts of the country and almost no deaths look very suspicious."
That is what you don't get, there was never an outbreak in the rest of the country outside of Hubei. The number of cases never exceeded 3,000 in any other province.
No, I'm saying that Iceland's numbers are strong, that they compare remotely well with most other nations (ie more testing, lower rate of infection), all of which, when taking together, make China's numbers look fabricated.
"That is what you don't get, there was never an outbreak in the rest of the country outside of Hubei."
3000 is an enormous number of cases, tantamount to a 'major outbreak'.
Wuhan's medical system was in complete crisis between lockdown and when the first field hospital opened 1.5w after: https://www.scmp.com/news/china/society/article/3047613/chin.... There's reports of entire families dying at home (https://www.theguardian.com/world/2020/feb/18/coronavirus-ki...) -- I am quite doubtful Lombardy has been hit harder than Wuhan.
I will admit I'm wrong if China runs serological tests on Wuhan to determine the percent that were infected. I'll be extremely surprised if they do because the data will prove the point above (my guess? it would show 1M+ were infected, meaning only 2500 deaths was impossible).
Death being concentrated in one region is not an exclusive phenomenon in China. If you look at Korea's stats you'll realize there many provinces with zero deaths[1]:
Incheon: 58 cases, 0 death Seoul: 426 cases, 0 death Gyeongnam 95 cases, 0 death Ulsan: 45 cases, 0 death Chungbok: 44 cases, 0 death Sejong 46 cases, 0 death Deajeon: 45 cases, 0 death Chungnam: 127 cases, 0 death Jeonbuk: 13 cases, 0 death Gwangju: 20 cases, 0 death Jeonam: 9 cases, 0 death Jeju: 9 cases, 0 death Airports: 202 cases, 0 death
Deaths are only concentrated in 5 regions:
Daegu: 6624 cases, 111 deaths Gyeongbuk: 1298 cases, 38 deaths Gyeonggi: 463 cases, 5 deaths, Busan: 118 cases, 3 deaths Gangwon: 36 cases, 1 death
[1]https://en.wikipedia.org/wiki/2020_coronavirus_pandemic_in_S...
NYT had a fascinating story on an effort in the US to scale up production of ventilators, that produced nothing in 14 years
https://www.nytimes.com/2020/03/29/business/coronavirus-us-v...
When you do widespread testing as China did in Zhejiang you can find a lot of mild cases. Indeed, Iceland, which did the same, has 1020 cases and 2 deaths.