I will first provide a link, and then I will go to your question specifically.
For discussion of the underlying mechanics I refer to my 2012 answer: gyroscopic precession. (That discussion, illustrated with diagrams, capitalizes on symmetry to explain the phenomenon.)
(Most attempts at explanation invoke the abstract concept of angular momentum vector. That rather defeats the purpose, because the angular momentum vector is itself a complicated concept.)
As discussed in the linked answer: the pitching motion occurs in response to rotation rate along the axis perpendicular to the pitching axis. That is: the applied torque causes a motion, and the gyroscopic effect occurs in response to that motion. As discussed in the linked answer: the pitching motion itself causes the gyro wheel to exert a torque. The magnitude of that responsive torque is proportional to the rate of gyroscopic precession.
As to what that means for engineering a device:
I expect that marine gyro stabilizers just have a large surplus of actuating capability, and only the rate of change of angle is measured. I expect that the software that controls the actuators is set up to act in rapid response, increasing the power output of the actuator to whatever value it takes to enforce the intended rate of motion.
Generally: the rate of gyroscopic precession elicits a corresponding torque. The elicited torque opposes the torque that led to the gyroscopic precession response.
For a gyro stabiliser: to keep producing the same effect the actuator must overpower the elicited torque. Assuming the spin rate of the gyro wheel will be high: at high rate of spin of the gyro wheel overpowering the elicited torque will be, by far, the dominant factor.