I decided to set up a more exact test set up this afternoon and I'm sure glad I did. The tank failed.
It was tested with a manometer set up to avoid over pressurizing. The water column is pushed up by the air and, when it reaches the pressure pre-determined by the length of the "U", simply blows out into an open circuit.
The first test was a 4 foot head, the height to the overflow plus a foot of margin for vent blockage. This is 1.73 psi, pretty minimal.
Here is the history of the pressure.
Going down pretty fast.
This is the bubble:
You can see a very short video of the bubble I made to send to the builder here:
http://www.cruisingonstrider.us/videos/Tankfail.wmv
This leak could probably be fixed like the others but I'm done. Triple M would have produced a whole new tank in the time I've let the builder screw around with this on. I'm especially concerned because it repeatedly seems to be tight initially and then develops another leak. I'm already out quite a bit of time and some real money for testing, fittings, running around, and I now need a new tank monitor sensor.
Even if it ended up tight, I would be worrying and thinking about this tank for the rest of my cruising days. I'm done. I'll be on Triple M Plastic's doorstep when the open Monday morning with this tank for them to use as a pattern.