Thanks for posting that wiring! I was just working it out on the breadboard, and I had come to the same layout as yours (after frying the first LED
However, I only have some diffused Red/Green & Red/Yellow bi-color LEDs on hand, and what I'm getting is a dead spot in the middle of the rotation. Both colors are lit, but so dim, there's no perceivable color at all.
Since everything else seems consistent with your layout, I'm led to believe that a) diffused vs. waterclear matters, or b) the colors matter. Just thought I'd post my results..
LOL I Almost fried my first LED but got lucky
The thing about this schematic, like you noticed, is that when you get to the middle of the turn, both LED's get so little power that the color gets very dim. In effect, all that a bi-color LED consists of is two LED's in the same piece of plastic. The problem with diffused is that they let less light through than a non-diffused. This can be good if you want six different very specific shades of red, because if the LED is tinted, it only lets one very small spectrum of light through, and the rest is reflected back by the plastic encasing and turned to heat (ever notice how a diffused LED can be easier to fry than a normal one?).
In Short, my setup has the same problem that yours does, but because your LED's are diffused, they don't let the whole range of light through, only one or two specific channels. This makes the problem much more noticeable.
If you notice in my picture when the pedal had a violet-colored LED, the pot isn't in the middle, it's more to the red side, because I have the same problem that you do: albeit mine is probably a bit less magnified.
It might be conceivable that the colors also matter to some degree, simply because some colors (Blue and Green) get brighter than other colors (Red and yellow).
There's definitely some experimentation to be had!!
Diodeates?? Wasn't he a Greek philosopher? An intellectual foe of Socrates, as I recall, because he could only think one way.