Thanks to Ozan for helping off-list with some additional measurement advice and insight on expected values. I continue to be amazed at what some of you can understand from a schematic.
He suggested replacing Q309/Q317 in the trigger pulse generator with the pair from my working 7B70, and that did the trick: stable waveform with sine/triangle/square, and controls work as expected.
I had previously tried replacing them individually with a 2N3906 (one of the few PNP transistors I have on hand), with no luck. Replacing both Q309 and Q317 with 2N3906s actually shows a waveform instead of a blank screen, but it stutters so I've ordered some 151-0220-00 replacements.
Adam