Probably a stupid suggestion!
My mobile phone stopped working up the hill today because of sub zero temperature, so I put it inside my jacket until it warmed up. Which made me wonder about possible transient temperature effects on the 787.
Now Li-ion is much less degraded by low temperatures than other chemistries, but it must still be temperature dependent to some extent. Assuming (!!) the charging voltage is adjusted down as the temperature drops, then it would also be adjusted back up as the temperature recovers on descent and landing. But not every area of every cell will be at the same (transient) temperature as the relevant sensor at all times. If this temperature discrepancy exceeded the built in "safety" margins, then over voltage could be applied in some local area(s), leading to damage, and failure becoming apparent after landing.
Of course, the reverse temperature changes occur on take off and climb, but maybe the margins are deliberately greater in that phase, to minimise the probability of an in flight problem.
And I know that I don't know what I'm talking about, so it is more of a question than a suggestion.