Final Control Test - "Fuel cut-off valve" causes RPMs to increase

I have a ANH-type engine (a marinized 140-5 5-Cylinder VW TDI) with some issues. After having a technician replace the old fuel injection pump and adjusting the timing, the engine is running but showing some problems - when the throttle is increased the RPMs increase as well, but once they reach about 1800RPM a little more throttle causes the RPMs to decrease back to 1200RPM.
As part of the diagnosis with VCDS the ECU values all look OK and are in range, but one odd effect is that the "Fuel cut-off valve - N109" test causes the engine to immediately go from idle to 2800RPM. When the throttle is moved forward from idle the RPMs stay at 2800RPM until the normal throttle position is reached for that speed and then increase higher. When the throttle is moved back to idle the RPMs stay high. Once the engine is shut off and started again is is the normal 775RPM idle with the same original problem.
There are no error codes shown.
The engine is hot right now and I'll let it cool down for a while, then I plan on

• checking the cut-off relay
• removing the cut-off switch and testing functionality
• seeing if there an obstruction in the line
• if the tests above are good I will see if I can find a wiring loom or connection problem

Can anyone think of a possible cause? If it isn't mechanical, then I'm wondering if the ECU might not be having serious problems?
p.s. The original thread for this issue is at Looking for expert sleuths - identify a VW (car) Engine
Additional information

• checked all the relays (removed them, applied 12V and tested resistance of the 87 and 30 leads). No problems or strange sounds
• Noticed that the resistance between the cut-off solenoid and the engine block was 0 with the single lead removed (that was odd to me)
• Applied 12V from the engine start battery to the single lead on the cutoff and heard the cutoff solenoid click each time

Additional information
I went through some more tests from the service manual and it would seem that the ECU has some problems, since the first step in the final control diagnosis is the "commencement of injection valve N108" and when that starts I hear a distinct change in the engine rhythm. Going down that diagnosis path had me check resistances, cable connections with the ECU removed, etc. and none of them failed, which leads to the final sentence in the manual that the ECU should be replaced.
I do have one question, Pins 1 and 2 in the ECU connector are shorted, I don't have a description of the pins but looking at some pictures on the internet it would seem that they are both +12V connections. Since the harness is the same as VW car ECU connections someone here might confirm that for me. The diagram is shown below:


Additional information - 2 days later
When the ignition is turned on, there is +12V to the fuel cut-off switch. When the VCDS runs the fuel cut-off test the +12V remain on that line instead of going to either 0V or to ground. I've checked the cables and there's no short to any other cable or to + or -. So my working assumption is now a fault with the ECU.

I forgot to update this thread with the explanation of what was happening:

The original fuel injection pump had a short-circuit in the quantity adjuster. Jesse from Banchwerks in Providence had a spare top-end from a VW TDI so we replaced it and used the hammer-mod to get the fuel quantity correct; but the engine ran a bit rough and would have the RPM issues.

So I ordered a rebuilt car injection pump, it had to come from Germany since the 5-cylinder ANH version engine was never sold here in North America. The new injection pump showed exactly the same problems.

After checking all lines, cables for resistance/shorts, and replacing the ECU, the fuel filter and even the fuel lift pump the problems remained.

It turns out that car engine cut-off solenoids need +12V to open. Marine engine cut-off solenoids are normally open and shut off when powered with +12V. So the car injection pump had an incorrect shut-off valve installed. It seems that the fuel pressure in the engine was strong enough to let the car solenoid pass a limited amount of fuel sufficient to get the engine running but insufficient for high diesel volume.

The old cut-off solenoid was installed but the problem remained! It then turned out that the diesel technicians in the Caribbean had neglected to reinstall the solenoid with the little spring that keeps it open! So without a spring it partially shut as well. Once the spring was installed and the QA hammered to the correct position the engine runs like a charm.