The last time I attempted to measure the barrel / cylinder misalignment, I was seeing 0.0015" misalignment using a tenon stub that was offset 0.005". From that I estimated that the true misalignment was 0.006"- 0.0065" and I cut the barrel tenon with 0.006" offset.
Today I made a tenon stub with 0.0065" offset (or another way to put it is 0.013" runout on the test indicator) and measured again.
-- 3 holes would take a 0.357"+ gage pin part way, and a 0.356"+ gage pin all the way
-- 3 holes would not take a 0.357" gage pin but would take a 0.356"+ gage pin
The 357"+ gage pin was still rubbing on the same side of the cylinder where it has always rubbed, approximately 3:30 o'clock as viewed from the muzzle.
Based on today's measurements, I currently estimate that the misalignment is 0.007" - 0.0075", so in hindsight my original estimate of 0.006" misalignment was not enough.
Also, when I roughed the barrel I had not anticipated needing to offset the tenon -- that idea came later -- so I had turned the diameter to 0.676". The tenon is 0.670" x 36 tpi, so that only left 0.003" of excess metal on either side of the bore to work with, assuming the roughed barrel was concentric with the bore. Well, when I went to finish the barrel, I realized that 0.003" of excess metal was not going to be enough to give me 0.006" offset.
The math did not add up! All I could do was going ahead and dial in the bore with 0.006" offset and hope for the best. In theory it might have worked, but in theory it might not have worked, depending on thread tolerances.
Fast forward to the most recent water tank test where some of the bullets showed significant misalignment. The explanation is 1) the intended 0.006" offset was not enough and 2) there was not enough excess meat on the roughed barrel to properly create 0.006" offset on the tenon, anyway.
So this barrel is unsatisfactory. No biggie since I only paid $15 for it years ago and I rescued it from the scrap bin for this project. Plus, I have one last mission for this barrel -- I've always wanted to try a Taylor throat and this barrel with its known misalignment should make an appropriate guinea pig. If the Taylor throat is a failure, that's no big deal because I have already decided this barrel is scrap.
Plan "A" is to recut this barrel with a Taylor throat.
Plan "A.2" is to cut yet another barrel, this time with the tenon offset 0.007". This is a budget project -- if I wanted to throw money at it it would make more sense to fabricate a line-indexed cylinder -- so purchasing a new barrel is not going to happen at this time. However, I have a Green Mountain barrel on the 357 TC Carbine. That barrel has been rode hard and put away wet, and has pitting and erosion near the throat, and I haven't shot it in years, so it is an eligible candidate for salvaging. I'll cut off the eroded portion and make a new barrel blank from the remainder.
Bear in mind that even if I cut the tenon offset just right -- 0.007" or whatever the true misalignment is -- there is still no scientific way to clock the offset tenon -- I simply mark the offset point and clock it by eyeball. Also, there seems to be 0.0005" or so variation from one cylinder hole to the next, so even if do everything right, there could still be 0.001" -- 0.0015" misalignment. This offset tenon trick is never going to result in perfect alignment, but the goal is to end up with better alignment than the factory setup.