You would be hard pressed to run the rotor RPM up to the point where it would start to come apart. Drag increase due to RPM will, at some point, become self-limiting.
You may do some long term damage to the components, and damage that might make it necessary to replace the dynamic components for no other reason that they haven't been tested or proven beyond a reasonable speed, but I doubt you'd ever overspeed anything to the point of immediate failure during your real autorotation.
The only catastrophic failure I know of related to speed was due to an engine runaway from idle (65% to 117% in two seconds) that severed the tail rotor driveshaft because of the sudden surge in torque (more than just speed).
Has anyone ever heard of an overspeed that was catastrophic?