Assuming you're using conductive plastic pots and not carbon, and that your pots are the right value and taper (and don't have anything weird going on like dead spots at the end of the resistive track), it's interesting that the issue has persisted despite changing the pot. Do both pots you've used test fine over their full rotation, out of circuit? What brand / model are they?
If all is well with the pots themselves, the next thing to look at would be the pot wiring - triple check all is correct at the pot end, and also check the wire-to-board connections too. I'd go so far as to check continuity with your multimeter between the pot terminals and the relevant solder pads of R42, R52 and U8B.
If all that checks out, next thing I'd do would be to test every component in the high shelf filter (page 4 of Audiox's schematic) with your multimeter against those in the working channel. Maybe there's a misplaced or duff component, a short, or a cold solder joint that this may help expose. The fact that you've got one working channel is gold for troubleshooting, as is the fact that this seems confined to a single filter section - check and test that part of the board rigourously and systematically, and you'll no doubt find the culprit. If you upload a couple of decent pics of that area of the misbehaving board (component side, and solder side) we can get some more eyes on the problem as well.