John,
Do you have a spare idle controller that you could use as a swap experiment? If the problem was thermal/mechanical, e.g. cracked vacuum line, I would not expect it to get better above 80C. Also, if the temperature sensors/wiring/inputs were bad, would not expect the behaviour to change across temperature. So, I was wondering if the controller is making bad decisions between 40 and 80C...

You could mimic the idle controller (don't remember if it is a constant voltage or a constant frequency) to fix it at 750RPM with the engine warm but not hot. Then, drive it and see if it still surges. This would verify the surge is not coming from sources other than the idle-stabilization circuit, also would eliminate the valve itself and related vacuum lines from suspicion.

Just some more ideas, Andrew Seidel.

__________________
86 560SEC
85 500SEC Lorinser
83 300D Turbo
75 240D