===> A GFCI, one which is built to NA standards as requied by law, monitors the difference in current between the hot and the neutral wires. Whenever that current varies too much, I think it's about ten milliamps, the thing trips and turns off the power. IMO, he was given a glib, and wrong, response by that person.
===> It would be fairly rare in any instance, but moreso as you indicated. By the time a spike makes it through the transformer on the pole, and all the inductance in that xfmr, there isn't much of it left for the house wiring, which in turn has a substantial sinking ability due to the things plugged into it. I'd be willing to bet money that no transient was at work in this picture; it just doesn't fit the anecdotal evidence.
Since the original
===> Now, that's a distinct possibility. In fact, I mentioned it in a post a few minutes back that I'd neglected to consider that. If that's the case, then there was an obvious reason for it tripping that anyone worth their salt could have asked about, instead of just that "they do that".
Umm, no, not really. High frequencies first of all wouldn't make it thru the network into the home wiring and if it was powerful enough to make it thru via arccing or whatever in the xfmr, it's not "noise". There is a lot of "noise" on any of those power lines; it's the design of the lines that protects it from getting into the houses. Else, anyone with a modem could pick off the billing information, all kinds of data that are also flowing up and down those wires. Any signals on those wires cancel themselves and anything left is further reduced by the turns ration in the transformers. There is even a reason for the spacing between power lines, in fact, that comes into play. You can only decode data AT th epower lines termination points. By design, any outside noise is also going to be horizontally applied, and will again be self-damped. It's called longitudinal balance. I used to work on that kind of equipment. It's interesting stuff, actually, since few people know it's there. Electrical and gas consumption billing are two typical applications and anything from lightning to the sun can induce noise into the lines, but due to design, it's reduced to the point where it has no afffect on things.
===> You cannot say that. The gfci tripped, and the reason isn't known. You can not know whether there would be a dangerous fault or not if the GFCI hadn't been there.
===> Like I said, that's possible, but very unlikely to be the problem, given the information available so far in this post. Since we're now going into empirical, anecdotal eveidence, I have a GFCI on my pool pump, too, because I also run the pool lights from it. Then that receptacle goes off to another receptacle/switch, which is used for the yard lighting. It has NEVER tripped before, during, or after a storm. The only time it's ever tripped was when I pressed the test switch, or stuck a resistor between the conduit and the receptacle. There are some valid reasons why some of the larger horsepower motors will cause them to trip, but it's not the noise generated. It's the pase shift caused by the inductance in the motors, especially when they're capacitor starts instead of clutch-started. Either way, the currents in the live/neutral vary sufficiently for the GFCI to detect the current difference, and thus it trips. I recall the last one I bought, a portable, had a reaction time in the micro-second range; that's pretty fast when you consider you're working with a period of what, about 16 mS on 60 Hz? I don't think any of that's relevant to the OP's situation.
However, I DO think your comment about something plugged in at the time of the rain or, whenever the thing opened, is important, since it wasn't specifically noted one way or the other by the OP. That installer sounds either incompetent or sort of, uhh, dishonest and doesn't care since he came up with "they do that".
Cheers,
PopS