I didn't read the full paper, but have you taken a stab at whether there are any GMs that do consistently outperform others by a standard deviation or more? If there are 60 people of roughly equal talent competing for 30 positions (which seems to be the assumption driving current salaries), then it doesn't much…