I wonder if some of these cases might be caused by the "phantom power" supply required for electret mics.
I was wondering when someone was going to bring that nonesense up. Answer is that's ridiculous.
To expand on that information, "phantom power" is supplied to a microphone over what is known as "balanced" audio lines. On the common three-pin connector on pro mics {known as "XLR"} the power is carried on pins 2 and 3 (+) and the negative (-) is on pin 1 (the shield of the cable.) It is a DC voltage, current limited, and anywhere from 12-48 volts. (48v being the most common.)
The normal connection scheme for balanced audio mics is Pin 1 = shield; Pin 2 = "signal hot" (or +); Pin 3 = "signal return" (or -). {Note that some brands of gear reverse the
signal polarity of 2 and 3}.
A "dynamic" (does not need phantom power) mic is connected between pins 2 and 3 and thus does not see any voltage across the element if the phantom is accidently left on. (Both pins are at + vDC potential).
Since the (-) end of the phantom supply
is tied to pin 1 , (and thus possibly to the chassis) it would be at the same reference as the chassis and the shield of the cable, and therefore the phantom could not cause a shock as described in the thread.
And if the (+) side of phantom shorts out to shield, the supply in virtually all gear shuts down to protect itself. (That's in case a mic where pins 3 and 1 are tied together in an "unbalanced" arrangement, which obviously creates a short across the supply.)
Hal brought up a great point, that the pools have heaters, filters, lights etc. any one of which could have energized the pool.
So a properly connected mic, with a mixer or amplifier in proper working order, may have actually provided a ground path in this case.
Like Hal said, we may never know....