I missed that. I saw the link, then started looking closely at the pictures, and forgot all about the video
I guess check orientation of components that can easily be reversed, like electrolytic caps, diodes and the JFET transistor. With some transistors there are variations on the order of the legs, for example.
Extra checking of all that off board wiring seems called for too. With all those wires, unintended feedback / coupling, wouldn't really surprise me. Maybe spread it all out on a bench and see if the behaviour changes when things are further apart. Hand effect is sometimes an indicator of that sort of problem - if the feedback or glitching changes when you put your hand near it, or a lump of metal, something may be coupling to something else, just by physical position.
An audio probe is a good start. If you get the datasheet for the 4013, it makes a lot more sense, but the idea is there's a circuit that turns your guitar signal into a square wave, involving two of the op amps used as comparators, then that signal clocks a divide by two circuit, made out of part of the 4013. I originally thought they made a sub octave and a sub-sub octave, like a Boss OC2, but they don't - there's just one octave down. The rest is a choice of filter settings.
They are actually quite hard to get to track properly anyway. The official advice was to turn the tone control down on your guitar - for less harmonics, and play single notes, which ideally means muting any strings you're not playing, as bass players tend to. Rest your thumb on the low E string to mute that (you won't need it - you'll never get that to track properly) and pluck a string with a finger. If you're playing a note on, say, the G string, rest your fingers lightly across the higher strings to mute those. Try plucking further from the bridge, to get less harmonics, and turn the tone knob way down.
Some of what it was doing did sound about right. Making a square wave out of your guitar signal makes for a very synthy sound, which then has to be filtered down to sound less synthy - and I did hear synthy. The sort of sound you had going into it was probably not going to track easily either. You need the guitar to sound less bluesy... ideally, you'd want it to sound something like a flute.
So, with the audio probe, see if you're getting a square wave signal on one pin of the 4013, and then one an octave down from that. IIRC, one side of the 4013 is part of the square wave production - I seem to remember a set/reset latch, driven by the two op amps feeding that chip. The other half of the 4013 is the divide by two - it flips its output once per complete cycle of the square wave fed into it, making one an octave lower. The signal from there goes to the gate of the JFET transistor, which should produce a signal at the drain.
In particular, look for a square wave, very synthy signal, at the gate of the JFET.
I guess the schematic I'm looking at is already in this thread. I seem to remember giving my take on how to get a useful sound out of one - which is not immediately obvious, IMHO.
The weirdness at extremes of the pots is definitely not right though.
Even so, maybe play with your signal going in and see what happens, because they are quite awkward things to deal with - even the off the shelf ones, like mine (now with just an added op amp output buffer).
Do you have a looper? If not, maybe an MP3 player or a tablet... or a phone, playing a simple clear tune would do - or a recording of your guitar. You want to try it with clean, dull sounding notes, and try it at different amplitudes. If you have any sort of EQ pedal or filter, use it as a low pass filter to soften the tone.
The thing with turning a signal into a square wave is that any zero crossing is seen as the same thing - a tiny crackle gets turned into a full volume square wave, just like a clean note does... or at least, it has that potential.