for instance, what if we're just interested in range-extension? can we transform the motion created by combustion into electrical power in a clever way? cylinder-solenoid coils?
are poppet valves so great? suppose we have some other mechanism to create the motion (solenoids?) or rotating valves? something electromagnetic appeals because it gives complete control over timing (rather than a crankshaft).
rotary engines are appealing, for the same contrarian reasons. but they seem to either have practical problems (wankel) or don't seem to be making it to market (peanut-shaped rotors, etc).
if 6 cycles makes sense (presumably in combustion physics), does it make any sense to burn in one chamber, then move those products to another chamber for some further (potentially different) cycle?
would it help if you could ignite from more than a single place? multiple plugs sounds like a bit of a pain, but could you generate an annular spark? would you want to control the location-timing of the combustion front? does rotating-detonation have any meaning in this context?
are there ways to reconsider the materials engineering of engines? make them dramatically cheaper, lighter? one of the best EV arguments is simplicity, but how much of current IC engineering is based on assumptions that can be broken?