I have been managing a 300kWatthour battery bank now for a couple of years and so am not a complete newbie when it comes to batteries, but am somewhat perplexed by the experience.
My own personal battery bank is colour coded red for positive and black for negative, arranged in switched banks, fused and arranged using with connecting wire and bus bars appropriate to expected currents, fuse ratings and out of preference I like stainless steel connecting bolts and washers.
One of the major reasons for failure in my experience of banks of traction batteries is the insulated bolts clamping down on a 1mm ring of metal resulting in a good recipe for bad contacts: I'm aware of these sorts of issues and for that reason in testing this bank I replaced them by stainless steel and ensuring good contact with connecting wires held securely by washers.
The first cell that blew was in a bank that someone asked me to look at. Its major problem had been poor connections. I charged it as one of the subsections of my battery bank on which the charging currents per section are not much more than a ball park of 12 to 15 amps maximum. For this connexion with 6mm cable is perfectly adequate. For discharge as a test load I have a 22V to 30V 600W grid tie inverter which from a 24V battery bank on a continuous basis puts a constant 500W into the grid, so discharge at 20 amps. No-one can say that a C50 rate is mis-use of a battery . . .
That discharge to 22V was a few weeks ago and from memory it was over 30 hours, thus 600ah
On the day that the cell blew, there was good sun, giving a 12 amp charging rate at around 28Volts which should be fine even if overcharged as an equalising charge . . . and as you say with the caps on, the hydrogen concentration should not have led to ignition, but there was ignition from some source inside the cell and the explosion was violent. The explosion was internal and not external.
With regard to the second cell which blew under a desulphation charge the specialist charger used, developed by someone in the battery business for over 20 years, does not use a high voltage pulsing regime. It so happened that the developer of that charger was here and it was he who had set up the three cells to be charged. The sources of ignition were internal rather than external and it's for that reason that this is of particular concern. Were an identifiable cause in terms of a loose wire, external spark to have caused an external explosion, those are comprehensible circumstances.