Does it really matter? Everybody agrees that smoking is bad, and I am sure that no members of a longevity forum smoke 
It’s also well known that you can’t just “transplant” findings from one country into another. The culture, habits, genetics, and behaviours of the people are also important. You’re not going to get Americans to behave like Hong Kongers or Swedes, no matter what government policy you manage to implement.
I don’t see anything special in any of these data. The biggest factors are increasing wealth, not being obese, and having access to healthcare. You can see clear examples of that within countries, such as the clear, large, life expectancy gap between poor and rich people across the USA, UK and most of Europe.
Look at the UK data (for men):
Bottom socioeconomic 10%; life expectancy 73.2y, healthy life years 49.8y
Top socioeconomic 10%; life expectancy 83.6y, healthy life years 69.2y
That is almost 20 YEARS more healthy life years, and 10 extra total years of life. So if you can raise up the poorest people in society, you make absolutely massive gains on average.
It isn’t simply a matter of universal health care. The UK system is pretty bad nowadays. But as I said earlier, culture and lifestyle have a lot to do with it. Even when the bottom 10% have just as much right to see a doctor as the top 10%, their jobs are probably shittier, worse workplaces, worse housing, worse childhood experiences and upbringing, higher psychological stresses (job insecurity, bills, debts, crime), poorer health literacy etc etc.
A much more obvious explanation IMO could simply be genetics. Hong Kong population is 92% ethnic Chinese. Japan and Korea are also highly genetically homogenous. All of those countries seem to be somewhat “protected” from negative effects of smoking compared to western populations. For lung cancer, typically smoking is a 3-5x risk enhancer in East Asia, but a 9-23x risk enhancer in the West. (pubmed.ncbi.nlm.nih.gov)
There are other situations where things like this are a major factor. For a 10mmHg increase in systolic blood pressure, a white person has a 8% higher risk of stroke, but a black person has a 24% higher risk. (https://jamanetwork.com/journals/jamainternalmedicine/fullarticle/1485077)
There are also clear differences in how populations respond to certain classes of drug. Clopidogrel is an anti-coagulant given to people often after a stent placement. It needs an enzyme (CYP2C19) to activate it. Around 60% of East Asians are poor metabolisers of the drug, and so compared to a European, East Asians taking this drug have a 2x higher risk of MACE, and a 5x higher risk of stent thrombosis. If you treat with a different drug (ticagrelor) then the risk is equal between East Asians and Europeans, showing that it’s the difference in drug metabolism which is the factor.
Long story short, IMO it’s likely that genetics explain a huge amount of the difference.