If it were my pedal board I'd redo all the wood piece extensions so that the first contact was being used... the MIDI Encoder board being used defines that first contact as MIDI note 036-C, the next as 037-C#, etc...
but it is using the first switch, the remainder is the end, the first photography tricks, it seems not, but when the pedal low coincides with the first switch.
In any case, the problem of C # is going backwards, is another problem, the transposition is already fixed.
Oh, Ok... I see it better now. The issue with your note On/Off for C# is caused by the magnet being too close to the reed switch and closing the circuit when the pedal is in neutral position. As the pedal moves down the magnet moves and the magnetic pull on the reed switch changes and allows the reed switch to open, thereby sending a MIDI Note Off message. As you release the pedal the magnetic pull again changes as the pedal goes up and the reed switch closes the contact and a MIDI Note On signal is then sent. The only solution to that situation is moving the position of the magnet. The problem is not the reed switch.
The second physical pedal (C #) until pressures will not sound, but once you step on when you release event for "on" and begins to play, and if you do not want to sound have to have it pressed, as if that switch worked in the opposite direction.
Moving the other pedals has no effect? It is possible that the magnets of the C and the D notes add extra pull, but only you can experiment with this and see what is actually happening. From your description it sounds like just the C# magnet needs to be reset....
There's a program called "Building Blocks", that I used for my PCOrgan many years ago. You could do anything you wanted with MIDI data. In your case you would put it through a "block" that would increase the MIDI note number by 1.
But I feel there must be something fundamentally wrong for this to be happening.
EDIT: You found your problem, so this post is irrelevant.