The specs required a highly mobile tank capable to destroy any Warsaw pact (PAVA) tanks at long range with a high hit probability on first shot. This led to the crafting of highly precise system.
To be honnest with you there is no stabilisation on the Leclerc. The gun is slave to the ballistic computer which computes the ideal LOF from the stabilised LOS.
When reloading, the gun goes to the reloading elevation. Meanwhile the LOS is still stabilised to the direction of observation (in the limits of the mirrors amplitude). Unless you release the palm switches, the mirrors go to their mechanical neutral positions.
The gunner sight is mechanically mounted to the main armament. When the gun goes up and down; the sight bows up and down.
Since the both move along with the exact same angle, boresighting can be done automatically with a deviation measurement laser (AMX 10 RC being the first french AFV to be equiped with such device).
Crews do some alignments (what we call "harmonisation" where we keep the parallax in check), but that's not the bullshit stated by Sergei Suvorov where crews were forced to boresight everytime they move their tanks...
At the time engineers were open minded on what could replace the classical tank. Once they defined that their platform was still an AFV, they assessed every kind of compromise to take what was the most favorable and compatible to their specs guideline.
Fun fact regarding the tracks. They spent quite some time to switch to steel tracks. They initially used the same arrangement as the aluminum alloy tracks (the shape of the rubber trackpads were supposed to reduce the stomping effect). Surprise, surprise, the vibrations at high speed were strong enough to be a handicap. This explains why we transition from V2 (alloy) to V5 (steel). Apparently V4 was also a disappointment.
Even with V5 or DST 840 the vibration is quite awkward compared to V2.