My take is that the author is focusing too much on "all factors", and the reason for the distribution is that the people who 1) realize they are "good" and 2) want to improve, end up putting in more training time to get even better, which leads to a natural gap in measured ability.

Consistency may be the largest factor in peak performance, whether physical or mental. Sure, consistency alone won't get you to the Olympics sprints if you're 4'11" or your neurons don't fire as fast as the next person's or you are prone to injury, but you certainly won't be setting the 100m world record (in the modern era) if you didn't already put in years of training, day after day, whether you were feeling 100% or not.

At the professional level, you get paired up with coaches, you get compensated (maybe), you are definitely obsessed enough to be training so hard. Of course there's a gap, and it makes me wonder just how linear the distribution would be if every able-bodied person trained maximally and effectively.