weird interaction between 2 pedals - can't find the reason

[R.G.]

Hmmm. The short answer is that your testing has almost completely confirmed the self-oscillation RF theories. It appears that SOMETHING in the setup likes to oscillate, and that it's fairly finely balanced to oscillate. The list of proposed fixes were all aimed at stopping self oscillation in some fashion. You've found that all (except changing the zener) actually do fix it.

As to which is best - we don't know exactly what is actually happening in there, nor what "best" would look like. So we can only guess on general principles. I'd vote for doing things at inputs rather than outputs, as things on outputs might only be filtering out high frequency RF that's still there, keeping the beast caged if you will. I actually like the 100R on the gate lead best, as it conforms best to my personal theory of why it's oscillating, but I'm prejudiced about that.  The input ferrite bead and either 47p input cap are likely to be as good in your situation.

Take your pick.

[intripped]

I've chosen the 100R res, and modified the cloned BOR accordingly.
As RG said, I also think that it's better to stop the oscillation at the (presumable) origin, instead of keeping the beast caged.

thanks to everyone for the help!!!

[Rob Strand]

I've chosen the 100R res, and modified the cloned BOR accordingly.

I'd probably go with the 100R and a cap to ground.  The cap to ground would help contain HF behaviour
of the circuit to a local area around the MOSFET.   This reduces the chance of external wiring creating a condition where the oscillation could start-up again.   

If you put in the cap and the 100R then it starts oscillating you already know the 100R isn't good enough because it's relying on the external conditions (wiring)  to stop it - and these could change with different pedals and cables etc.