Well the right tank has been on test since Thursday and I was keen to see if it had held pressure or not.
Using the digital manometer I checked the pressure and was delighted to see that, after 36 hours, the tank pressure had risen! This happens because the outside air pressure has risen causing an increase of pressure in the tank.
I wrote the results on the tank like I did all the others just for the sake of completeness.
That, I think, means the right tank has passed the test! Yay!
Next I tackled fixing the left tank leak. It is leaking along the top baffle to skin joint. After some considerable thought and discussion with several people I decided to go with the vacuum method. This involves connecting a vacuum pump (I found a cheap brake bleeder on Amazon) to the vent tube and sucking the air out of the tank until I was near 1 PSI of negative pressure.
Then I mixed up some Class A PR1440 sealant which is much more liquidy than Class B sealant and ran a bead along the entire join and rivet heads.
After a short while I could see that the vacuum was working and sucking sealant into the gaps where there were leaks.
I applied more sealant here and restored the vacuum to 1 PSI again. I repeated this step several times until it seemed to stop creating these gaps and holding negative pressure much longer.
Once that was done I put sealant on the seams and rivets of the right tank too just in case in case. Then I finally put sealant on the access plates of both tanks and screwed that in place for the final time. Now I need to leave the tanks for a few days for the sealant to cure.