Over-running clutches can fundamentally slip, and re-engage. Seems this application has had more than it's fair share. Maybe the metal had issues, but Problem here is the sudden torque spike from the other engine goes through a drive shaft that usually is not loaded much, and could actually be loaded in the opposite direction. This spike could be big enough to bust things, but looks like in the June 2022 case anyway, it caused the remaining side to also have the hard clutch re-engagement. This would never happen in any conventional design where the engines power a gearbox that then powers the rotors. By keeping the engines outboard, that shaft is doomed to experience loads that include this impact- and it appears that the design did not address this load. It is a fundamental catastrophic failure mode that should be predicted and mitigated. Maybe future tiltrotor designs would be better to have inboard engines, and just tilt the prop gearbox.