6206b update:
The voltage range switching works. The supply is overspecified though, so that it, unloaded, will be able to go pretty far over the limit of the lower range. Fools you.
The meter selector / range switching assembly was luckily screwed together, not riveted. Two screws hold the decks together in expected fashion, and I was able to separate them out enough that the severed press-fit between first aluminium deck and brass bushing could be freed from the assembly and subjected to the arbor press. I located it so it would go together and just two fingers on the press lever was enough to push it together again. Built up a jig with a staircase block from the drill press clamping kit underneath, and a 1/4" socket (IIRC 8mm) as press interfacing bushing on top. That was the most pressing (sic) problem I knew of, because a broken switch of course would require a parts donor.
Anyway, once that was taken care of, I started bolting the chassis together again. Before replacing the lid, I subjected the supply to a series of tests within the capacity of my lab. Mostly, I'm lacking in resistive loads, so had to make do with a lightbulb for load.
This of course means I can't do all tests I want to, but the ones I can do indicate no problems. The meter was zeroed and once that was done, the metering was found to be in calibration. Of course it is non-linear in overvoltage.
Ripple is one order of magnitude below limit in manual.
Meters used for the verification include the MetraHit 25S, the Fluke 8060, and the -hp- 427, 428b and 3438 meters.
Conclusion: Supply is ready for the bench, once I've done a Med style full power burn-in. As soon as I have a suitable load resistor I'll get on with it.