Why would a parallel group short to itself if the voltage is the same across all packs? no voltage difference, no current transfer..
The problem is usually in serial connections, where voltages are differing.
Shorts will, 99% of the time, melt the connector, wires, solder, or anything else in the pack before it actually does any kind of damage to the pack. This happens instantly.
There will never be an accurate analysis of what happened with his pack, it seems like there was not abuse, so i can only assume that bad new pack syndrome was to blame. Perhaps one was puffing, and the more it was charged, the more it got aggravated..
If you have a dud in 2.4kWh, the likelihood of a massive fireball is really, really high.
~2.4 kWh will be made up of approximately 24 6 cell 5AH packs.
That's 144 cells total.
And all it takes is 1 cell to light the whole thing ablaze..
this is why i am constantly beating people over the head to test their packs fully before even using them on a quality RC charger, such as the iCharger, or Hyperion series. Those tests involve a full discharge down to 3.0-3.3 while logging the cell voltages using software like the included logview. If you see a cell with a major anomaly, you don't use that pack, period. If a cell is misshaped or puffy from the factory, you don't use that either, period.