With apologies, I model physical systems for my job. What we do is take measurements and do the math and use the math to understand the measurements. And I know that most posts that start out this way are BS. And I don't think the following is a particularly good way to think about penetration, but it does explain the carbon arrow youtube comparisons that make light arrows look better than they are.
I have integrated Newtons second law from the surface of an animal to where it reaches zero velocity - the penetration distance - for different forms of the resisting force.
The results are different for friction - such as slowing down an arrow by squeezing on the shaft - and for "plowing" resistance - such as a boat moving through water.
This is friction in the physics sense - friction forces oppose the motion and are independent of the velocity. For frictional forces the correct variable to calculate penetration is KE.
For plowing forces, forces which increase in strength with velocity, the correct variable is momentum.
So if I shoot a light carbon arrow and a heavier aluminum arrow into the same target, since friction is the primary force opposing penetration, the carbon arrow is not penalized for having less momentum. But it does benefit from having a smaller shaft such that the target squeezes on it less hard and the frictional force is much less. It penetrates much farther and a new industry is spawned and a million deer are lost due to lack of penetration.
For an arrow penetrating a deer, we do everything we can to minimize friction. I believe that the other forces that oppose the arrow's movement are proportional to the velocity of the arrow. For example, the arrow must move the flesh out of the way of the arrow - the faster the arrow is moving the more violently the flesh is thrown aside and the higher the force to do this. The calculus says that for forces proportional to the velocity the correct variable for predicting penetration is momentum.
Thinking about it, the light arrow is moving faster so the force opposing it is higher (since the assumed force is proportional to velocity) - the light arrow is penalized for its higher speed. Consequently the light arrow loses more energy per inch of penetration than does the heavier arrow.
Now things get weird. If the forces slowing down the arrow were all proportional to the velocity the arrow would never stop! As the velocity went to zero, the forces would go to zero, and the arrow would continue to move forever, although too slowly to see after a while. Of course there is always some friction so the arrow does stop - at least in the dirt.
And, for the sake of full disclosure, the fact that "plowing" forces are proportional to the velocity is from a web site that talks about modeling boats plowing through water (for game simulations. This kind of explains the last paragraph - a big boat really does take "forever" to slow to a halt.
F = ma = change in momentum per unit time
impulse = F times delta time = change in momentum
It is almost dawn.