After some probing we knew that lug 1 of all the pots connect to ground and that each control increases its own effect as the pot is rotated clockwise. The original used 5K linear pots for all the controls, and so if we use 10K for all the pots instead (also replacing the existing pots), keep lug 3 common and just switch the wipers, then we know that the entire value of the pot out of circuit will always be in parallel with the pot in circuit and so we get a predictable affect irrespective of the position of the out of circuit pot. 2 x 10K is the 5K used in the original. This will alter the taper and so reverse log will be the best choice [see discussion below].
For indication with the 4th pole I've shown a bi-colour LED with a common anode so that you will get a different colour for each "channel" selected. If you're not bothered about LED indication then of course a 3PDT stomp or switch could be used instead.