1) Herd immunity has kicked in, meaning a majority of people have already been infected and are recovering/recovered
2) Containment measures were effective against a highly communicative virus which can live in the air for 3 hours and on surfaces for up to 12 hours.
3) They just stopped testing and/or are lying about the results and are hiding the resulting health care burden.
IMO the only one here that doesn’t require competency on the part of the government is #1, making it by far the most likely scenario.
If #1 is true to any extent, then it means IFR is actually quite low, more like 2009’s H1N1.
The other fundamental question I keep going back and forth on is the final total population infection rate. For H1N1 and other flus it’s usually on the order of 10-30% of the population. (There are arguments why this one might be higher, although I think some people confuse exposure numbers with infection numbers...)
Containment supposedly will not meaningfully reduce this ultimate number but would just slow down how long it takes for everyone to ultimately be exposed.
So while there are very few new cases in China (and they are closing their temporary hospitals) what happens when people go back to work?
Cultural factors: how quickly do people change their habits, to do e.g. social distancing. Do people in public wear surgical masks. Population density, temperature, weather,...
Lying governments: I don't think South Korea or Italy are lying, but Iran is for sure, and China up to some point (at least early on). Italy stopped testing at some point, but this is known. South Korea increased testing over time.
Herd immunity: maybe a part of the population was already immune before (too early to say), but it couldn't have "kicked in" during this year as there was only one wave yet.
There is no question whatsoever that this corona virus IFR is a lot higher than the 2009 H1N1 flu, but lower than the 1918 H1N1 flu. When comparing countries / regions, you should consider the population age pyramid (Italy has more old people than South Korea) and how overwhelmed the hospitals are (I think quite bad in Italy and Wuhan).
Shouting this point does more bad than good. Outbreaks are exponential in their initial stage, to an extremely good level of approximation. Sure, they're ultimately s-shaped, but we're far from the bend point. That the growth starts to deviate from the exponent a tiny bit when half of your country is sick isn't a useful observation right now.