This has not been my experience of the unit, but much of the difference in their operation may be down to the very different usage that they get between our cars.
In mine, it is certainly disengaging, but that is under cornering loads and with a manual transmission and up to 12" slicks. It may not be under such constant loading as with an auto, but that may also be because the speeds where slip begins to occur is already above the 25mph stock 'engagement/ disengagement' maximum speed, and if much below that speed a gear change would be needed before very long.
No doubt you will also be starting off from full throttle from 0mph onwards and not lifting off at all, where on a circuit the throttle could be rolled on and off or feathered to the suit available grip and cornering radius so the load fluctuates.
I'm more prone to stop, spin a wheel just enough to lock, then proceed over/through the obstacle, that way I avoid walking home
The first step for me then is still to increase the engagement speed a little by reduction of the big mass.
The ferocity of engagement is a valid point, but even so I believe the shock loading on the half shafts and across the diff forging itself is the real issue. The clutch friction plates will only see this very briefly until the force on the ramps locks it all up when there will be no significant further slip.
Last time I went shopping for one, I saw a couple with the clutch ears all twisted off.
It would be interesting to see if any of the shock load could be reduced by softening of the springs controlling the rotating weights to reduce the required half shaft speed difference that activates engagement?
I'm curious as to the cause of failure of the G80 in the photo. Was this diff already modified to allow engagement at higher speed.... it looks unmodifieed in the photo? Did it fail on initial engagement at low speed or when moving faster? I don't personally believe that either the G80's clutch packs or its ramps will be the weakest link in the chain.
I don't think I've seen any evidence of the ramps failing first. The one pictured is from a 4wd with knobby tires, one spins, then hooks up and the case fails, or it doesn't have the grip to fully lock, and the clutches burn.
The fracture of the main diff body forging looks like a material failure around the big mass pivot pin bores? I take it that it failed during actual lock-up?
it locks, or it dies trying, I think breakage once locked is rare.
If serious wheelspin was to occur on a high-output car running big slicks and a modified G80 at higher road speed and rpm, the shock torque load across the axle immediately as the diff locked could be huge. There might even be an argument for reducing the clutch pack friction area and introducing more slip during lock-up to try to 'soften' the engagement?
Seems to me there isn't enough clutch materiel to support prolonged slip Eaton makes nice clutch diffs, and this isn't one of them. If I got 1/4 second of slip, I bet my 7000 lb wagon wouldn't snap sideways half up the freeway onramp by my house every time it rains.
We can't really complain too much about a unit failing like that that if we are asking it to do much more than it was ever designed or intended to, but for some uses it could still be a useful and low cost option with only minor mods.