As RR NDB has stated, It's NOT just the voltage that's critical, but the current as well .
Maybe you thought (previous post) I was talking about "Toy" helicopters?....
Think about a 90 cm rotor-disc (six foot in old money), that can,and has caused fatal injuries, and you'll appreciate that these people are pushing Lithium batteries to their limits.
Motors get so hot,the field-magnets actually demagnetise.!!!! Battery-packs commonly give 5-15 minutes of intensive aerobatic flight, their life is limited to ~100 cycles in this very demanding application.
Battery self-destruct is an accepted risk ,simply because of the high energy-density and massive discharge-rate capability.
These batteries [B]need[B] individual cell-monitoring on both charge and discharge..they are now well-enough understood andthe monitoring technology well enough established, that thousands, if not millions of serious hobbyists are willing to put a month's wages in the air, powered by Lithium technology.
Direct connection to a bus is NOT an option...charge/discharge MUST be through appropriate monitoring -regulators.
Perhaps it's a problem with getting approval for an unorthodox connection-interface?
Total demand is irrelevant, the total capacity of ALL the battery-reserves on the 787 is only a few minutes of full demand.
I'm sure that properly monitored and conservatively rated, these batteries could become acceptably safe....effective controllers will keep them fully charged and prevent an unsafe discharge regime.
I strongly suspect the control interface (charge/discharge) is the primary culprit.
Agree with LYMAN. dendrites destroying a cell are probably caused by defective cell/battery management.