Here's a quick followup.
I thought the problem looked like an issue with the machine resetting, and thus re-starting the power up routine over and over. The bar graph would progress farther and farther until it finally made it through, after about 30-60 minutes. It looked like something was becoming stable as it warmed up.
I could not locate schematics, and when I opened up the unit, I was disappointed to see the part numbers have been abraded off the chips. However, I did notice several bulging electrolytics.
I applied some heat with a heat gun in the region of the electrolytics, and found I could quickly get through the test and into the operation mode before a reboot, but still, it was rebooting. Freeze spray took it back to rebooting during test.
The most bulged cap was a 1000uF electrolytic, so I replaced all four of the 1000uF with new caps (and a 470uF soldered in parallel with one of them on the back side), and apparently, that was sufficient to get it to boot normally. The self test now takes only a second or so.
I'll replace all the bulgy caps on the main board. Everything looks OK on the analog board, so I'll leave those alone for now. Then I'm sending it off for calibration.
I'm not a subscriber of the recap first and debug later approach, but in this case, it was indeed bad caps.
Dave