Greg, introducing paraffin oil in place of some castor reduces the viscosity in a controlled and limited way , reducing ether by substituting it with kerosene leads to an uncontrolled reduction of viscosity. This reduction can be dramatic since the paraffin oil is about a sae 15-20, while kerosene is sae 1.
What I am searching to imagine is what is the physical dynamics of the fuel droplets in their travel from carburator jet to the head. It could be interesting to prelevate some sample of the liquid phase in the crankase (two strokes I mean). There we should find the oil-kerosene mix that wets the conrods and the crankshaft. But, how this mix is a true mix? Over a certain percentage of kerosene the two components within the droplets should separate. Forces within a droplet are particularly high, so the oil could be embedded within a mantle of kerosene.. This could mean that inside the crankase we we have no more a solution, but a particular dispersion of (castor I mean) oil in kerosene, allowing for an adequate lubrification. The castor could stick to internal surfaces better than kerosene. This could explain why mineral oil behaves worse than castor in diesels... and this could be an argument against olive oil, that mantains kerosene in solution more easily than castor.
Ugo