I dont know if its a condition of when in "rotary" the majority of the work is being done by the outer diameter of the pad while the center has a much slower surface speed and really isnt doing much, and the outside of the pad gets used up faster while any compound on the inside just lays there.
Given that the working motion of the tool while in "rotary" mode with the adapter isn't as pure as the rotary would deliver to a directly driven pad, I'd have to think this makes sense.
The pad locks into place off-center, which causes its rotation diameter to be larger than the diameter of the pad itself. At the outer edge of this motion the pad foam would be forced to compress and "dig in" to the surface in an unnatural manner, which presumably would cause the center portion of the pad to lift away from the paint AND would heat the outer edges of the pad more than the center of the pad.
In orbital mode, there is still an off-center motion occurring, but the diameter of each stroke is significantly smaller, allowing the pad to compress and "dig in" evenly, involving the entire face in the work of buffing as you noted.
Sure, that makes the rotary mode effective for cutting purposes, but it certainly isn't suitable for final polishing and will cause premature wear to your pads if used for long periods of time. I'd say the Dynabrade's rotary mode is probably best suited to isolated repair operations as opposed to full-vehicle buff-outs when paired with foam pads. I'd imagine that the behavior of a wool/foamed wool/surbuf type pad would be somewhat different and would deliver a different result. Regardless, it's fairly easy to remove the adapter and swap in a regular backing plate to use the rotary as a true rotary.
I'll try to get around to doing some testing within the next few days and see if there's a way to capture the pad motion on film... maybe running the buffer on a piece of glass and recording from the underside.
:buffing: