Jonathan, Which weighs more so which is the more logical frame of reference? This isn't something that I dreamed up. All of my various references on the topic frame their view of damper function this way. It is more a matter of definition than anything else. Everyone talking about it has to use the same reference frame or there will be mass confusion.
The reason that the sprung mass oscillates after a speed bump is because the unsprung mass has been allowed to impart some or most of it's inertia into the sprung mass. Prior to that imparting of inertial energy the sprung mass' inertia had it going flat and level. Notice that it is delayed, the sprung mass' movement only starts to happen after the the tire is past the bump. It isn't the bump compressing the suspension that upset the sprung mass, it's the suspension unloading. So it is not the compression damping that is at fault, it is the rebound damping that is at fault. Or at least is the cause. It is extremely easy to blame the compression damping because that is the direction that you are moved when experiencing such an event, but if you train yourself to look at where that upward motion happens to the sprung mass relative to where the bump is you'll see that the tires are past the bump, so how could the initial compression have caused the sprung mass' reaction?
If the dampers could do a perfect job (slightly over-damped) then that wouldn't happen. But because that perfect damping would only be effective for that size bump encountered at that exact speed there is no point in making it perfect there because it will be imperfect everywhere else. The old "test" for shocks that need replacing is to jump on the bumper and if it only goes up once past ride height and then settles out then they're deemed "good" by that test. That is a perfect example of Critical Damping. Which is what most strive for. In a vehicle that means that some of the inertial energy in the unsprung mass from hitting a bump will be transferred to the sprung mass - upsetting it's perfectly flat & level traversal of the terrain. Put in electronics terms we have a variable frequency, variable amplitude input and we want Critical Damping for it. That's a pretty sophisticated damping function. Velocity sensitive damping is pretty close. If you can add-in position sensitive damping you'll gain that last 10% or so (my guess). Since adding that in is expensive it fails my cost/benefit analysis for a regularly driven vehicle. A race vehicle is a different story.
Having used deflective disc valved dampers for all sorts of types of driving over the last 15+ years I feel that they are the best choice within the U.S. for any purpose. Not that there aren't other good choices too, but on a cost and availability stance there isn't much that can compete.