The engineers were complaining that the monitoring they were doing couldn't react fast enough due to slow computers to stop the overrun in time to save the engine and turbopump.
I think this is the writeup:
SSME history in several parts
From part III:
If only the normal operating torque were applied to the HPOTP without the fluid load applied (gas in the pump or in cavitation) it could accelerate from a dead stop to a destructive overspeed condition in less than a tenth of a second. The acceleration rate under this condition is almost 400,000 rpm per second.
And probably unnecessary at the technology level they're trying to achieve. They could easily produce pressure fed or gas pumped designs which might not achieve the pinnacle of mass ratio, but would have the decided quality of being more reliable. Even under duress of bad management decisions. I noticed the nuke programs (eg the Paki one too) had similar rigidity of design selection. Maybe the management culture doesn't reward innovation. Perhaps they might be more successful with an upper management change...?
Heheh...