After the 48D failure, for the following nine launches, a 4.25 second holddown delay was incorporated in order to verify engine stability before liftoff. Also a second redundant accelerometer was added to the Rough Combustion Cutoff system to increase reliability in case an accelerometer would fail. No booster instabilities were detected in the nine following flight although in the attempted launch of 32D the RCC in the sustainer engine was triggered, the engine was subsequently replaced and 32D had a successful flight. Of course 27E would show that the problem still existed.
I guess the 48D footage would resemble that of 9C which also burned on the pad for some time before exploding. Of course this was during a Flight Readiness Firing and not an actual launch.
9C also produced an incredibly powerful explosion that leveled the entire service tower on LC-12
I read the postflight report for 48D and it gives a fairly detailed description of the events during the attempted launch.
51D actually should not have lifted from the pad, but the RCC sensor in the B-1 engine was not working so it allowed the missile to be released anyway. The B-1 RCC sensor on 48D wasn't working either but the malfunction occurred in the B-2 so that engine's RCC operated correctly and terminated thrust before lifoff could be achieved.
Interestingly enough, Big Joe, an unmanned Mercury prototype, made it through MAX-Q even though, the Atlas failed to stage. No escape tower on that Mercury-Atlas launch either.
Ultimately the blame for the failure lay on Flight Director Walter Williams both for ordering the launch to take place without the LES and for launching it into cloudy weather where the booster couldn't be filmed after the first 20 or so seconds of flight (as well as depriving us of footage of what was probably a pretty cool explosion).
Also on the topic of MA-3, I can think of at least three other flights that went straight up and didn't pitch over. These were a pair of Thor launches from VAFB in 1959 and an Atlas R/V test in 1968. The Thors failed because some technician forgot to cut a wire holding the programmer tape in place so they never executed the pitch and roll sequence. The Atlas (95F specifically, it had an ABRES TVX vehicle) I'm not sure of the exact reason for the failure, but probably similar circumstances to MA-3.