I don’t own one either but if you’ll check older posts in this forum you’ll find many complaints and a couple solutions for buffer (feeder/pusher) issues. I don’t recall any surface finish issues.
Did you rerun input shaper tune after tensioning the belts? I run it at Z=25, Z=75, and Z=150. Large variance with increased gantry height indicates a structural issue.
Long shot but implement the PLR patch (PLR and ‘mcu’ shutdown: Move queue overflow). The existing PLR performs a file system write inside the motion planner. It is possible (but very unlikely) that is causing micro jitter on following the Z move.
Then try using “paint on seams” in the slicer. Paint them right down the middle of the bad side. This forces the printer to Stop in the middle and may give insight to resonance issues. Corner to corner could be informative too.
Lastly manage your expectations. Positional error is motor steps PLUS deflection. Deflection is calculated as a percentage of the length of the build axis. All things being equal a bigger printer can never match the precision of a smaller one.
Partial solution to the above is print slower… and even slower. Once you eliminate resonance effects and get baseline THEN start working on resonance prevention.