Seems to me you are looking at the question sideways.

On any chamber there is an optimum spark advance. Going over that number makes less power. So it is wrong to assume that more advance makes more power in all cases.

What you want is a combustion chamber that makes maximum power with the least advance. That cuts down on negative work (piston pushing against expanding mixture).

If you are able to find the maximum power spark advance on your engine with regular cooling, then by going to reverse cooling you could run more compression.

If you were to put the engine on a dyno and keep increasing spark timing until it started detonating, without reaching a maximum point in the curve, then running reverse cooling would allow more advance that would increase power.

R.