Okay, at least that blog post is written by someone with a phd, but you can't really take this at face value either. It's probably reasonably true in general (obviously still highly dependent on the person doing the running), but you seem to be assuming that this rate is the same on short runs and long runs, which sound suspicious. I wouldn't at all be surprised if the rate of energy expenditure increased as the time/distance of the race increased. Also if they're running instead of sleeping, I imagine that is probably increasing energy use as well.
I'd also be pretty surprised it the 100 calories/mile was taking into account the person's resting metabolic rate. That could easily add another one to two thousand (more depending on the person) calories per day.
Finally, even with your analysis coming to 1.5lbs of fat, that's quite a lot of fat in a trained athlete. I'd expect most of the men running these races are well below 10% body fat, and the lower your body fat is the more resistant the body tends to be when it comes to losing more. Losing 1.5 or more lbs of fat would probably be moving their body fat percentages down quite noticeably, which is not easy to do at such low levels.
So maybe it is all bs, but I don't think it can be ignored so easily either.