Rewired the lambda sensor wires after that had melted on exhaust. Used the rest of the screened cable and as an extra precaution used some earthed braided pipe to cover from the loom to the screened cable.
Then took her to the garage and back. I knew immediately that the sensor was working as she shot off like a scalded cat. Had to work hard to stay below 60 mph.
Next thing I noticed was the temperature gauge started working. So it appeared to need rewiring after all even if it hadn't melted on the exhaust.
Finally, she behaved perfectly after the fuel stop, which historically is when the lambda sensor stopped working.
A quick check of the data stream showed:
The data from 0-1.4 mins was when I ran the engine in the garage to check the wiring. Start up time was less than a minute. 1.4 mins - 8.0 mins was the journey to the garage. The start up time was 1 min and she was good from then on. At the garage the start up time was only 30 secs and then she was on song again. I can see there was four sluggish areas for a few seconds on the way back but they were not noticeable in the car and certainly nothing like the drop outs I was getting before. Interestingly they were almost exactly a minute apart. How do we explain that ?
Later: Figured it out. They all coincided with me lifting my foot off the throttle. Turns out then when you do that the ECU stops the fuel completely so the engine slows down quickly. The lambda sensor senses a very lean mixture and drops to 0V. So it is working as it should.
Time to take a few days off I think. The fittings for the new heat shield come early next week so that is the next project.