Very interesting, I'd never considered how much current might flow in the ground relative to the wire, though I was aware it existed. Doing a little napkin algebra (very iffy early in the morning!), the ratio of the total wire resistance to total ground resistance is proportional to the square of the line length. Using your numbers, a short single phase distribution line of only a couple of miles won't have much ground current because the total parallel ground resistance is still high compared to the wire, and most current is still carried in the neutral. Which is good for me because that's how our area is wired.