Lets say you had:

A Briggs Stratten overhead valve engined single cylinder 3000 watt generator that runs at a steady 3600 rpm

Two 1500 watt electric heaters to create a load

Two Kil-A-Watt true electrical power meters able to measure up to 1750 watts and accumulated kilowatt-hrs

http://www.amazon.com/P3-International-P4400-Electricity-Monitor/dp/B00009MDBU

How would you run an experiment with this generator like a mini dyno laboratory
With the single spark plug indexed different positions
To see if indexing
makes the slightest difference?

What type/brand of sparkplug?

Should the recomended gap be used
or something bigger or smaller?

Would the effect of indexing be seen more
With a tiny gap?

My first thought is to measure fuel efficiency,
but would measuring detonation resistance be better?
Faster combustion is supposed to improve detonation resistance at the same octane.

87 AKI octane gasoline
could be diluted with diesel fuel
to find a lower octane blend at which the engine would begin to ping at full load,
kinda like what they do with pure Heptane and Octane liquid blends at a State Gasoline lab with the special CFR engine.

What thinks you?