[Jamesp]

Many years ago in engineering school I had a fuels and lubes class with a lab. We used a knock engine to determine RON (Research Octane Number) of various hydrocarbons including gasoline straight from the pump and various gasoline brands and octane ratings blended together. The result was generally the RON in any brand was understated at the pump, and blending various grades and brands occasionally resulted in an octane higher than the highest grade in the blend. The professor (a retired API wonk) speculated that different types of octane boosters in each brand were present in slight excess to guarantee the stated octane rating, so mixing the brands allowed the excess booster to raise the octane on the other brands by using a different chemistry to boost octane. No idea if it's true, but I've never had a problem mixing brands and grades due to this experience.