... the surge protection equipment in the substations detects this, and `stalls` the supply for a second,...
There is no way anything can "detect" that surge and "disconnect from it". Before any mechanism even starts to move, the surge is way long time gone.
The only way to really suppress the surges are the parallel (to the line, in other words from the line to ground) arresters, either in the form of the "discharge fingers" around the insulators (the first line of defense; absorbs most of the lighting energy) and their faster, but less robust semiconductor (like a VDR or so) "buddy" at the substation (it is way faster in response, but can not handle as high energies, it is supposed to handle only the first rising peak the slower "insulator finger" still passes). With all of these we are talking about response time in microseconds (the complete lighting strike takes just few 100 microseconds before it is over)
But the problem with these is, once they are ionized, they force the voltage down (that is the main reason why they are there), but then conduct the electricity as long as they are fed by the current. And because their drop uses to be lower than the operating voltage on that line, the current could be fed from the power network. This will obviously destroy them.
So in the substation this is detected (in fact it is detected as any other ground fault or an overcurrent) and that line is temporarily disconnected from the power for some second or so and then connected back to restore the service. This happens way after the overvoltage event, but the aim is not to stop any overvoltage or so at all, but to break the current feed to the arresters, so just extinguish the arcs there and restore them back into non conductive state.
So the lineman was right about where the lighting current flows, about the automatic shutting down the line temporarily, but wrong about the purpose of that interruption...