Hello Royaloak,
Skating force varies between 4 and 18% of the VTF, depending upon all sorts of factors. Any compensation will result in a compromise. My suggestion is such a compromise, reducing the sidethrust/lateral preload on the cantilever and, at the same time, retaining an amount of skating compensation that will be sufficient to keep uneven stylus/groove wear at a VERY low level.
The often described method of using the tracking ability("torture") track on a test record will lead to overcompensation due to the fact that there is no 70µ, 80µ or higher constant modulation to be found on any record. Peaks: yes! Permanent signal: not supposed to exceed 50µ, typically around 20µ.
Since modulation causes drag, it will also "modulate" skating, so why compensate for a maximum that rarely, if ever, occurs.
Overcompensation will often cause cartridge dampers to assume an asymmetrical position after a while(cantilever no longer parallel with the cartridge body/generator axis). Once you notice that, you ought to send it back to the manufacturer to have it re-aligned. The stylus(when kept clean) usually outlasts the damper/its proper orientation.
Try a dynamic Mono(i.e. Sonny Rollins - Saxophone Colossus) record and play the right channel only, then the left channel only. With the antiskating set as high as you mentioned (moving slowly towards the outer edge of the record), you are likely to notice one channel to sound more dynamic than the other. Reduce skating and listen again...
All the best,
Frank