If a fault occurs within the circuit, on the conduit or at the end device, it will travel via the neutral/EGC back to Xfmr - this represents the vast majority of faults--But when a fault hits the earth or grid remote from the EGC, it wont simply disappear into the earth - it will go back to the Xfmr.
When a fault occurs to normally non current carrying paths that are conductive, that current travels via the EGC back to the Grounded Conductor at the transformer if wired correctly, and yes, this represents one type of fault condition.
Again. A lightning strike is not a fault.
I've never seen a transformer or overcurrent protection device ahead of a lightning strike....... have you?
Yes, a lightning strike does in fact simply disappear into the earth.
You can't stop a lightning strike by any amount of ground rods or grids.
If what you were saying were true, about a lightning strike traveling back up the GEC to the Grounded Conductor on a transformer in close proximity, lightning would destroy a ton of transformers each time one decided to strike the earth.
That's simply not the case.
JAP>