Possible scenario: The dwell (4 cyl. scale, right?) on solid 15° could be the ECU trying to lean out a rich condition. The O2 sensor bouncing around 0.5 volts says the ECU effort is working (and that the sensor is OK)

When you disconnected the sensor and grounded the lead (faking a lean signal to ECU), I think you said the rpms increased? (some punctuation and a return key now & then would make reading easier for these old lookers). Anyway, did you happen to see what the dwell did, as the rpms picked up? I'm guessing it moved up off the "pinned" 15°.

I'd like to see what the sensor reads when disconnected ("Open Loop"). Again, my guess is that it would be up around 0.8volts or more, and steady. (It only swings when reacting to the ECU's mixture adjustments.)

I adjusted my daughter's AMM ('87 240) that way recently, with a new, spliced in sensor, and it passed the Massachusets Emissions way good.

Of course you would have to get the plug out of the AMM, by drilling two holes and grabbing it with narrow-nosed pliers.

Well, maybe this post will get the ball rolling, and bring in some good tips.
--
Bruce Young,
'93 940-NA (current)
'80 GLE V8 (Sold 5/03)
'83 Turbo 245
'76 244 (lasted only 255,000 miles)
73 142 (98K)
'71 144 (track modified--crusher bound)
New 144 from '67 to '78
Used '62 122 from '63 to '67