'Net relative to other course of action' is the tricky bit. I find it hard to believe that investing in 15 people at $15 million, or 150 at a million, or however you want to break it out, would not be a better alternative.
The issue of CEO pay boils down to how skewed you view the distribution of talent among a population and how much a single person makes a difference in an organization.
Personally I think success is more of a function of collective or distributed intelligence, and perhaps if a single person is so pivotal compared to nearly everyone else in an organization that is more of a reflection of poor organizational structure that is not distributing responsibility and power sensibly. At the very least that sort of corporate structure could hardly be viewed as a robust system.
Thats not criticize this particular case, but highly skewed compensation in general.