Tonight I ran through most of the 519 adjustment procedure. I had noted two problems, namely that there were some odd trace artifacts on 2 ns/cm, specifically the trace was "curling up" on itself on the right side of the sweep. Additionally, the time base adjustment was not even close.
The trace artifacts were solved by adjusting V331, the 4CX250F that is used as the Miller runup tube in the 519 design. I found that not only was the control grid voltage off (this was my doing when I was trying to resolve this problem last month), the plate voltage was set 15V high. After re-adjusting for 155V on the plate and -3.2V on the grid, the curling trace problem disappeared.
I ran through the time base adjustments using my Type 184 as well as my 8664A synthesized signal generator (for adjusting the fastest time base settings). This brought everything into spec more or less.
Next problem to figure out is that it appears there is a dc short between the center conductor and the outer conductor of the rate generator output. The rate generator works fine, of course, but this makes it damned hard to inject a dc bias for say, testing diode reverse recovery times (which is one of the suggested applications!)