The equation for "KE Mode" works like this:
cost = (.5 * mass * deltavmagnitude) / 4
where deltavmagnitude = final velocity magnitude^2 - initial velocity magnitude^2
The major problem I have with this equation conceptually is it's clear bias towards a set frame of reference (namely, the observer).
I don't think you can calculate the force needed to accelerate an already moving object like this because the object may or may not be moving from any given frame of reference.
Assuming the object is alone in the universe (as indeed the equation supposes, since it includes no other forces) it has no way of determining how fast it's moving, or if it's moving at all. It's only when we take into account the velocity relative to the big blue screen that this makes any sense at all. And then you have drag and friction which sap the speed, not the velocity relative to it.
So the reason a car takes increasingly more Horsepower to accelerate faster and faster isn't because it's going faster and faster. If a car could theoretically work in an empty vacuum, it can continue accelerating at a constant rate given a constant energy input. The only caveat is that as it approaches the speed of light, relativity will begin to warp its sense of distance and time when compared to our "resting" frame of reference.
Back on topic, it's drag and friction and other resistive forces that eventually will lower a car's acceleration with constant horsepower to 0.
So once we come up with a realistic drag control setting, I see no reason to keep KE mode or attempt to fix it. Trying to 100 .up store will eventually reach a constant speed (if the slider is set right this can be below the max speed) where a constant 100 nrg is required to maintain said speed.
Thus moving at lower speeds is better because the resistive forces are disproportionately lower.
I think that's more realistic, solves the same problems, and works all around better.
