Using peak carrier velocity and two buffer weights, excluding and ignoring all other frictional forces(drag/hammer, etc.), you obtain the same or very similar KE between a +5.3 oz buffer and an 8.5 oz buffer(plus bcg @20 oz) with only a 6.5% reduction in peak carrier velocity alone. Carrier velocity is not constant in the cycle, the contributing factors are spring(weight) and buffer(weight) differences, if you add buffer and it slows the peak carrier speed down the release of the PE and the creation of KE is less(less felt recoil). There are also naturally occurring speed changes in the cycle of the reciprocating mass when the gun fires(to keep this simple just ignore that). The buffers job is to buffer the initial force on the reciprocating mass, and to slow it and give the spring a chance to absorb and control, then return. The full weight of the reciprocating mass should be considered(bcg & buffer). Full mass ar10 BCG's can vary 2-3 oz from brand to brand(5.3oz could quickly become 8.5 oz and really at that point its not a heavy buffer as its making up for a light carrier possibly).
The other acting force is PE and that relates to the spring, but for this lets say the spring stays the same in both comparisons with two buffer weights. If the carrier is moving faster it has more energy and carries that further rearward where the spring is storing more energy; going slower it travels less and stores less. Hooks law, force needed to compress scales linearly.
If you throw a racket ball against a wall at 50 mph its return force will be X, if you throw it at 75 mph it will be more than X.
With 20 oz carrier figured in: