Daniel Indiviglio poses the problem of pricing in-flight Wi-Fi. This is part of a class of problems in which fixed costs are high but marginal costs are low. In the economics literature, it is known as the "Disneyland Dilemma," from a classic article by Walter Oi.
What the airlines will want to do is try some form of price discrimination. They want the power user to pay a high fixed cost and low marginal cost. Thus, you may have a "frequent Wi-Flyer" plan, where if you pay, say, $40 a month, you can have unlimited in-flight Wi-Fi. The airline wants the low-frequency user to pay something, so they might pay $5 for half an hour.
I am not saying that this particular approach will work. But the idea is to come up with a scheme that maximizes revenue from all sorts of users. If you don't offer the volume discount, then the power user who takes four flights a month may just pay for half an hour of Wi-Fi per flight, and you only get $20 a month from them instead of $40. On the other hand, if you only offer the $40-a-month plan, you don't get any revenue from the folks who would pay $5 to be able to check email a couple times.
Anyway, it's a complex problem, but my prediction is that there will be multiple pricing plans.