Second of all, doing 1 test, in 1 zone, is hardly enough to even come close to "proving the claim". There are so many different issues at hand that could alter your results it's not even funny. Not to mention the fact that even if it was 100% accurate, it's still only 1 zone. If you had done the exact same thing to EVERY zone that had alliance players. And then made accounts so you could do the exact same thing to EVERY zone that had horde players it would be more convincing. I mean, you do realize that when you're alliance, Horde players won't show up in the /who list but they are counted for the total people online right? And same with the opposite.
That's just not true... you just need to test 1 zone. You run /who in the specific zone and get 250 people and then run your own test (of course, if test returns correct value) and get 150 people... you just proved there are faked players. Why would you have to test all zones ? You just need to test 1 zone to prove, that there are fake players. That's because when you are testing hypotesis you could support it with thousands of examples, but you need just ONE counter-example to disprove it. Yes, /who displays only alliance character, but count horde + alliance to total count (which is not BLIZZLIKE, fix it nos !), but I doubt there would be 100 horde players in elwynn forest. Furthermore, maybe his program count enemy players as well.
And about that 100 yard radius... player normal speed is 7y/s, which means you run 100 yards in about 14.3 seconds. Run for 14.3 sec and you will see how far you get, then imagine a circle with that radius. There are 11 of those circles in elwynn forest, yes it doesnt cover all elwynn forest, but it covers a lot of popular places. I could do a map if you'd like, with 11 circles with radius of 100y and centers in places, where characters stood in the video, to show you which area is covered and which not.