True. Summer fuel and non-oxygenated fuel, however, do have more energy than oxygenated winter-blend RFG, for matching octane levels.

Higher octane fuel has lower energy content not so much because of the refining level or length of the HC chains, but dominantly because of the additives used to boost octane- MTBE, ETBE, or ethanol. All reduce the energy content of a gallon (and oxygenate it). Yes, there may be more actual isooctane/less heptane in premium, but much of the octane gain is from additives. TEPb was used for the same reason (in addition to lubrication)- higher octane without higher refining cost (and subsequent yield reduction). Lower-octane fuel actually has shorter chains on average- dominated by heptane (7 x Carbon). 100% octane (8 x Carbon) is defined as 100 octane fuel. They're piped off the cracking tower at different points- same amount of refining, just a different boiling point.

But if an engine is designed on premium, then chances (in my experience of owning 3 cars that stated "91 or above") are good that MPG will suffer with lower octane. Not because of energy, but because of combustion factors- earlier ignition and whatnot the engine contols can't adjust for completely. One would think that any modern engine computer could handle any octane by now, but, being an engineer, I know 1st hand how easy it is to paint oneself into a corner, design-wise. Sometimes it is a whole lot easier and cheaper to get outself of the corner by putting limits or conditions on the end user. Ideal, no. Good for getting the product out the door, yes. Life is not an ideal system.