There should be an opposite thruster for each axis. I wonder if the short bursts were due to heating limits.
The short bursts are just the period of the control cycles. Control cycle starts, loop sees error, commands thrust; next control cycle starts, loop sees error is nulled (or in deadband), commands no thrust.
There are opposed thrusters, but I assume that in atmosphere and under parachute canopy it’s harder to make sure they are perfectly opposed.
Heating likely plays a role as well.
I am not a rocket engineer, but I have read How Apollo Flew to the Moon and Ignition!: an informal history of liquid rocket propellants, both of which cover these issues. Highly recommended.