Thanks for clarifying the fail safe setting. That it was set to kill the engine, rules out determining that voltage was lost due to the engine dying. The engine then could have died do to fail safe (loss of signal) or loss of RX power.
I've read many others say they set fail safe to kill the engine, but I favor an idle setting for the reason that if signal is lost and then regained, the engine is still running to bring the plane back. For the same reason, I don't set a fail safe condition on the electronic ignition switch. Of course, if power is lost to the RX, the electronic ignition switch will kill the engine but for loss of signal, it will only kill if fail safe is set to do so. Again, I favor letting throttle fail safe handle the chore of taking the engine to idle to await a hopeful return of signal.
It might be that during the early days of 2.4 when a reboot sometimes took several seconds after a loss of signal, that killing the engine on loss of signal was prudent... but rebooting is instant now for many if not most receivers and having the engine still running seems more reasonable.
Last, you didn't say if you were using dual switches but if only one was being used, the failure to fire up you described might have been a precursor of a switch failure and many switch failures start by becoming balky and not readily evident during the after crash investigation.