It was the cold winters in Tokyo
According to the NTSB report (and yes, I have read it all...) the reason Japanese airliners were the most affected was due to the cold weather in Tokyo.
The report states that Li-Ion batteries are more likely to suffer thermal runaway as a result of internal degradation caused by self-heating during high-current (charge or discharge) operations.
The colder the battery, the more likely self-heating is, and the more it occurs.
The NTSB concluded that the JAL and ANA airliners were more likely to suffer from self-heating because the crew climbed aboard after an overnight cold-soak in 0 degrees Fahrenheit and started up the APU. Full load. The APU then immediately started to re-charge the battery (at 45 amps...).
The temperature rise generated by that was suspected of raising the battery above the 146 degrees at which thermal runaway can begin.
The report seems to suggest that had the battery been 30-odd degrees warmer when they started up, no problem would have occurred.
It's a long read, but the suggestion is that these batteries are more likely to have problems if they are charged or discharged when they are very cold. In the report, they state (and EASA also states in its comments...) that the cause was never proven to either organisation's full satisfaction.
The report was very critical of manufacturing sloppiness by the battery maker, and oversight inadequacies by the subcontractor, the airplane manufacturer, and the FAA.