It's starting to make sense. The problem isn't one side getting dragged down, it's that one side ISN'T getting attenuated when it should be. Per the schematic, leads 7/9 & 8/10 are essentially common on the volume pot, with leads 9/10 being attached to a 15k resistor when the "loudness" switch is off, & that resistor being bypassed when "loudness" is on. If lead 10, as in this case, is open then the signal isn't getting attenuated as it should be with the "loudness" button in the "off" position.
Unfortunately, this still means the pot needs repair.