They mention they have a 20,6mm slave and a 19mm master in that thread.
Did you measure how much the pivot point for the arm moved when You swapped to the longer ball nut and added two washers under it? I moved mine about 10 mm towards the engine, no chance to get the clutch working with much less.
Unfortunately this was done a long time ago and my memory doesn't serve ! However i know that the clutch fork almost touched the M90's hole end that the fork comes through. I thought of doing that due to the fact that i wanted to be safe, but as it looks i was marginally there. I had also tried 3 washers but it wouldn't fit.
So, i removed the master cylinder today and measured the pedal's movement at the mounting point of the MC and i found out that from end to end is 32mm maybe 33mm. I could grind of course the upper stopping point of the pedal assembly a bit to have my pedal a bit higher if needed and gain some more travel, but i have to get the new MC first and then check this out.
It looks that the "lighter pedal - 0,7 bore MC" goal can not be achieved in my current setup, since it needs 32mm pedal travel, which means that it will engage-disengage right at the firewall which is no good.
So my best bet is this MC:
https://www.wilwood.com/MasterCylinders/MasterCylinderProd?itemno=260-6089
It has a stroke of 34,8mm - bore 19,05mm , dimension wise looks the same, i just need to figure out how to mount it on my pedal, since it does not come with an adapter or something. It has good chance of success since i need around 28mm pedal travel for the clutch to disengage out of the 32mm. And then i can check if i will grind the pedal assembly stop to lift the engaging point a bit further away from the firewall.
Another find was that my MC rod was a bit bent, which means that the pressure of these clutches is probably a bit too much for this pedal assembly. Maybe i got to reinforce the assembly like Dave Barton has done in the past.