Just in case Deejay is out of touch for a few days Ozzie, I'll pass on my suggestion for something you might try. It goes beyond what has been discussed before, so please regard it as an experiment.
There have been quite a few cases where captive-cotter clutch-halves have come to grief, ending up with the clutch bore loose on the PTO shaft, and the cotter moving further and further into the boss as the bore expanded. When dismantled these clutch halves seem to show signs of the clutch-end of the PTO key being chewed up. As Julius Sumner-Miller used to say, "Why is it so?" My new theory on that subject is that the (IMO rather dumb) design of the captive cotter system did not really clamp the key into position. With normal engine vibration the key tended to slide slowly away from the tightest part of the bore, which was directly opposite the cotter. So, the key found its way to the end of the PTO shaft at the thrust bearing end of the bore. It then encountered the thrust bearing. The key got beaten and mangled, and eventually stretched and damaged the bore of the clutch. Even before the bore was stretched, the eccentric forces on the key may have caused quite a bit of vibration, and fatigued the engine mounting deck.
Now, if this radical and completely untested theory is correct, you might be able to alleviate the vibration in your mower just by driving the key back towards the engine end of the keyway. Furthermore if you tried this and told us the result, we might have some evidence as to whether the radical theory is actually correct, or alternatively, like most radical theories it is complete moose-droppings.
