I think there were 2 main "tuck" eras, though the first was the only true problem:
1. P38 era, where the formation of shock waves on the main wing, at Mcrit of about .65 or so. When the shock wave is strong enough (depends on the airfoil shape) you get separation of the airflow over the wing, causing a sharp reduction of wing lift. That's a first contribution to the tuck problem. Secondly, the reduction of lift means less downwash produced, less downwash on the elevator, meaning the elevator is no longer producing as much down force (remember, the tail always pushes down, except for FBW fighters). These two were enough in some cases to totally prevent recovery from dives once a high enough speed was attained, even with full aft elevator. I seem to call some test pilots killed troubleshooting the behaviour. The only option was to reduce thrust, and perhaps add drag, to reduce airspeed.
2. DC8/707 and onward era. the following I'm not 100 % sure of, but I understand the tuck problem was solved, however, if you were to take one of these to mach .90 plus, the elevator was found to be "blanked". Ie again, full aft elevator would not recover the aircraft. But no problem, since they had a trimable stab, and trim would allow recovery. What I'm not so sure of is why the loss of downwash is no longer an issue with a trimable stab. I'm thinking the flow over the horizontal tail itself of the DC8 at mach .90 plus becomes supersonic, developing shock waves ahead of the elevator, and thus rendering the elevator ineffective. Thus trim can recover the aircraft.
These days we have mach trimmers, which function well before tuck onset. This is because the mach trim actuates to ensure the stick gradient of so many lbs of forward force required per knot of airspeed increase. You'd have to look up FAR 25 to get the number.
AFAIK, airbus and 777 (due to FBW) are exempt.
Perhaps some gurus can comment on the trimmable stab implications to mach tuck.