I think one of the modes of vibration is rocking of the axles through the pivot. In this mode the tires are acting as springs, and the entire assembly is rocking as a spring-mass-damper system, with, as noted, probably not a lot of damping. This system has similar characteristics across the pivot point, i.e. similar axles and tires on the two rear axles. When this this assembly gets to rocking, the motion tends to be perpetuating since the reaction of one tire feeds into the reaction of the other. When trying to suppress vibration in a symetrical system, one tactic is to introduce some sort of asymetry. So for my M925A1 with the 14.00R20 Goodyear AT-2As, I run different pressures on the rear two axles, say 45 front and 35 rear, and I think this has helped. I would be interested if anyone else sees any improvement from doing this.