Alcohol can be used to boost power because it refrigerates the intake charge, allowing a greater mass to enter the fixed displacement volume. It is suggested that the same can be done with ordinary hydrocarbon fuels -- in either of two ways.
(1) by selecting components having higher heats of vaporisation, thus causing more charge cooling at a correct mixture strength. Aromatics tend to have higher heats of V than most straight-chain or branched hydrocarbons, and the lighter fractions also have this property. Large amounts of such light fractions are used in some racing gasolines. Perhaps the entire intent is not just to obtain their good volatility?
(2) by running rich with fuels that are rich-operation-tolerant (as alcohol is). This approach simply exposes a greater mass of fuel to the evaporation process, thereby obtaining more charge density. Further, with sensitive fuels, the octane rating will be quite a bit higher when running rich, and this may be taken advantage of by use of higher compression ratio. I haven't seen this scheme work on the dyno, but people are definitely interested in it, and there is at least one fuel blender who claims his fuel works in this way to produce more power.