The persistence of the "big" bubbles, I had a hard time accepting but it leaves little other options than there indeed being a leak of sorts, and that would be for me at least, the first time to see that in the solenoid/housing combo.
If that is truly the case, I will have to review the way I fit the side-tubes. No way around that.
Vibrations make microbubbles, not big bubbles. Microbubbles CAN collate into larger bubbles, but NOT in this rate... That would typically be one 2 mm bubble every 10 minutes or so, IF there is a loop in the ful tubing where the bubbles can collect. The way you show it, with the fuel outlet of the solenoid pointing upwards, and STILL big bubbles, can't be anything else than a leak.

We have to figure out which part is leaking so I can replace the proper part.

One thing is sure, your engines CAN run on this system, I myself run engines with much smaller fuel consumption so that really is not the issue.

Meanwhile, what is your impression WRT lubrication? I have so far very limited experience with bushed engines (only an OS FP10) but I like to collect as much data as possible, because there are a lot of those bushed sports engines out there.
Do you notice a difference in the quantity of oil coming from the shaft?

EDIT: as an afterthought: you mentioned pneumatic testing. My request is to test under negative pressure, not positive. Sometimes, leaks along seals can have the character of a non-return valve. It is not unthinkable, that the leak opens up under negative pressure, but seals itself under positive pressure.