Am I missing something here? Aren't these results basically useless beyond trivia unless you normalize for the characteristic of routes. If an airline does mostly short haul flights, which are obviously more expensive by mile, then any graph proving that a short haul carrier is expensive seems to just show a tautology
In theory, the "fixed cost" should be captured by the $50 term.
Very cool. You should also graph them next to things like safety and customer service scores and other indicators of quality - and give overall rankings.