On the day of the crash, many of the witnesses were drawn to look at the plane because of popping noises (i.e., the engine noises was unusual). This was followed by many seeing flames between the wing and body (the engine re-igniting?). The crew did not loose control until they commanded throttle up for full power. If one engine was having fuel delivery problems, it may not have put out the required thrust so the plane then slew to one side and that's when the rudder let go (though many witness said they saw parts of the wing hit the rudder after the bang and flames).
As for sabotage, there are easier ways to damage an engine than by cutting through hard titanium tubing to a system that can only deploy with the help of the hydraulics. The above thrust reverser theory doesn't hold water.
The pods (nacelles) are designed to handle compressor stalls and engine surges (which don't destroy the engine, only cause damage internally to the engine, and often thrust is nearly fully recoverable by the restarting of the engine) without damage to the nacelle structure. If what you opine is correct (and it is certainly a plausible hypothesis) that one or both engines stalled or surged (and emitted a bang, and then restarted with some flame out the nozzle), this could explain the reports of explosions and flames.
Your other comment buried in the previous one concerned compressor blade failures. When this happens (either to the fan blades, the low rotor compressor blades, or the high rotor compressor blades), the engine essentially blows apart. If the fan blades depart, the engine and nacelle are designed to contain the fan blades within the nacelle structure, but the nacelle and engine are badly damaged and scrapped. More rare is when the high rotor compressor blades fail - nothing stops them except gravity and the ground, and the engine is irreparably damaged.
In both these cases, the aircraft loses the power in the affected engine, resulting in a yaw (the nose turns) toward the failed engine (since the other engine continues to produce thrust). If there was a stall followed by a restart, there should have been some indication of it on the flight data recorder, and the pilots would have commented on it on AA587. I don't recall hearing about that on the recorder data.
It is possible that the engines stalled (and banged, and flamed out) after the tail came off, and the recorders stopped running (10-12 seconds before impact) due to the high yaw angle of the flow entering the engine inlet. It would be interesting to know when the witnesses saw the explosions - before the tail came off, or after. If the latter, then your hypothesis may be a good explanation for the witness reports.