I was looking at the "I don't think constant *g* is" in the Recent Comments sidebar when that occurred to me. Even cases where it's training are dealt with - just treat the payoff as distributed over the period of training, I suppose.

Glad you liked it. I decided on two posts in case some people didn't know integrals or weren't concerned with the derivation of the equation and were willing to take my word that it made sense.

That's a good thought - sort of an inverse of what EY did with technical appendicies.