One thing the article didn't mention, that would be interesting to consider is the inflation rate of tuition from the time he was 17 until the time he graduated

One thing the article didn't mention, that would be interesting to consider is the inflation rate of tuition from the time he was 17 until the time he graduated. I'm estimating based on other research that it was likely between 8%-10%. That means the cost of his education would be roughly 50% more than he originally expected. Factor to multiply cost is 2.2 =(1.09^9) Figure it was increasing at a steady state so you could split the difference. Even if he had looked at salaries originally, then backed in to the price using normal (not education) inflation, he would have dramatically underestimated his exit cost.