Not to be critical of you in any way, but IMHO that is not a valid scientific basis for this requirement. Indeed, it is my belief that there is no valid scientific basis! I would be happy to see it removed. Consider this:
Case #1:
Run a 20 amp circuit 150 feet, using #12 wire and a #12 EGC. Postulate a fault at the load end. There will be some amount of fault current.
Case #2:
Run another 20 amp circuit 150 feet, using #10 wire and a #12 EGC. Postulate a fault at the load end. The amount of fault current will be more than in Case #1.
Case #3:
Run another 20 amp circuit 150 feet, using #10 wire and a #10 EGC. Postulate a fault at the load end. The amount of fault current will be more than in Case #1, and more than in Case #2 as well.
Discussion:
If we know that the fault current is higher in Case #2 than in Case #1, and therefore the chances of tripping the breaker are greater, why do we need to increase the fault current even more, by making the EGC bigger, as in Case #3? What makes anyone think that the fault current in Case #2 was not enough, given that the installation of Case #1 was legal, and that its fault current was enough by itself?
I don?t get it. :-?