You are right, Peter. I get by with pump gas because I have a mildly tuned engine and it is amazingly tolerant of fuel quality. A true high performance engine is less tolerant and race gas is the correct thing to use. This post is modified to recommend race gas instead of pump. This is how I use reference gasoline and a dyno to tune the mixture for Bonneville.
Step 1) The bike is filled with reference gas and put on a dyno. A good fresh racing gas is best. Always us the same type. In my case the dyno is at elevation 150 feet, more or less. Bonneville is at a much higher elevation and a lean mixture near sea level will be a richer mixture on the salt. I make a good guess based on the altitude difference and assuming the racing gas has a higher oxygen content than the pump gas I am using. The mixture is set at 13.5 to 1. See 2007 curve on attached.
Step 2) The bike is run on the salt. I check some mixture indicators. There are many and my preferences are the exhaust header pipe color, piston crown color, and how the engine sounds and feels under acceleration and at full throttle. In this case the pipes are blue four to six inches beyond the header clamps, the pistons are a nice nipple brown, there is no misfiring, and the bike runs clean.
Step 3) The mixture is adjusted as needed on the salt to get it right. I did not need to do this.
Step 4) The bike is filled with reference gas and put on the dyno. The mixture is recorded. The mixture trace from Step 1, if the initial guess is right, or the mixture trace from Step 4, if the mixture was adjusted on the salt, is the basis for the target mixture.
Step 5) The engine parts are examined during tear down to make sure these mixture assumptions are correct.
The head was ported in 2008 and larger intake valves were installed. Before we left for the salt, the bike was filled with reference gas, put on a dyno, and the mixture set as close as possible to 13.5 to 1. About 14 to 1 was the best we could do. See the 2008 curve on the attached. The bike ran great and the indicators showed a good mixture. Based on this, I reset the target mixture at 14 to 1.
This year I set the mixture at 14 to 1 on the dyno with the new build and the old CV carbs. Then I switched to flat slide carbs, guessed at the jetting, and ran the bike on the salt. It ran OK, but the pipes showed more color than I like and the engine did not pull hard. The bike was filled with reference gas when we got home and it was put on a dyno. The mixture trace shows that the mixture was lean. See the 2010 curve. In hindsight, before I left for the salt I should have filled the tank with reference gas, put the bike on a dyno, and set the mixture to my 14 to 1 target. It is likely I would have had a richer mixture and few more horsepower if I would have done this.