Through some serial kindness (Ed Milbourn had given a convergence assembly to Bob Galanter, who only needed the connector, and gave the rest to me), I've been able to fix the blue horz convergence.
Before doing anything else, I measured the convergence coil resistances. Turns out the blue coil is 2.4 ohms, just like mine, and the R and G are 3.0 ohms, just like mine. So, it seemed SAMS was wrong in listing all three as 3 ohms - probably as close as they could get at the time.
I went back to the set and:
1) pulled the convergence plug - blue H lines still drooping - so I guessed it's not due to >too much< blue H current. Then I put the current probe on the R, G, and B in turn and noted that all three had about the same H parabola current.
2) with the current probe on blue, started turning H tilt and H amp, noting as before that the tilt had a noticeable effect, but the amp had little or none. The H amp adjustment is a coil, and if you look way back in the thread you will find that I had broken the core and used a replacement from John Folsom. Well, the replacement has a much longer screw thread, so I decided I would just turn it as far as I could CCW and see what happened. Well, after a number of turns, the core started to really move out of the coil, and the current went way up - eventually I could get the blue lines to curve >up< on the ends. Readjusting the tilt (which was now wrong) and amplitude iteratively got the results you see. R and G current is about 0.5 A p-p, but blue needs to be about 1.1 A p-p. I was just too timid on adjusting that core.
All 0.2 A per div:
red
green
blue
result