The Shepard’s capsule seemed to be purely ballistic. Though I suspect there is a little bit of intentional misalignment of the thrustline and CG to help make it veer off a bit to one side when it fires, to make sure the capsule got away some distance horizontally from the booster. Apollo did this by using a separate small rocket motor near the top, firing horizontally to pitch it over a few degrees. Mercury did it by simply having the escape motor thrust slightly misaligned to cause a bit off-axis thrust line.
Anyway, it is had to tell when the Shepard’s Capsule abort motor actually burned out versus simply having some “afterburn”.
OK, I did some checking around. The escape motor burns for 1.8 to 2 seconds. The abort motor DOES have thrust vector control but it only works during that short thrusting phase. The wobble and tumble were well after it stopped thrusting. If it also was trying to use capsule RCS thrusters, they didn’t hold it. There’s no sense in purposely having 90 degree yaw to one side, and 90 to the other, and then a tumble-flip that seemed random for neutral stability. So if they were trying to use RCS as well to keep it oriented, they need to work in that.
Here’s a NASA Apollo Status Report Number 2, 1966. It has a lot of neat things such as Service module engine static testing, and RCS thrusters being tested. Also, three Apollo abort tests, Little Joe Mission A-003, A-004, and Pad Abort PA-2, beginning at a bit past 7 minutes into the video. Most relevant to this thread, is A-004 which did a pitch-up maneuver then fired the escape system. The front end of the Escape rocket had deployable canards that helped to flip the capsule around to make it point heat-shield first into the airflow. But it tumbled around a few times, FAST. Survivable, just not a fun ride.
[video=youtube;pvW03OYJ_1E]https://www.youtube.com/watch?v=pvW03OYJ_1E[/video]
Also note in the Pad Abort test (8:30 mark of the video) where that pitch motor I referred to fires at launch to push it from vertical towards a diagonal path. You can also see the nose canards deploy and start to flip the capsule before the LES separates.
In the A-003 Joe flight (before the 8 minute point) that broke up, the pitch motor was unable to make it pitch away because the vehicle was rolling so fast (this roll also prevented the canard system from flipping the CM around, but fortunately the drogues eventually pointed it the right way).
A-004 flight begins at 9:05. At about 9:34, the Joe is pitching up when the abort system fires. The capsule and LES tumble several times end over end at high speed. That was a somewhat worst-case abort as far as introducing multiple tumbles and high G-loads in various directions as it tumbled at very high speed and the air decelerated it.
Of course, lots of other neat stuff in that video, but the three Apollo aborts are somewhat relevant to this thread. Especially that fast tumbling on A-004