Early EDM non-cat 4A-GE 16v Bluetop Fails emissions (high co)

[Rascal21]

Pretty sure the W-pin is correct, in lieu of other info I`m starting to think ECU is switching earth for W lamp (while the lamp gets constant 12v).
If the pin-out`s were wrong from the Early 4AGE pinout available online the ECU would not be getting IGN signal (From memory, 12v IGN on pins 7-14 and B+ on pin 6 whereas W is pin 13).
And you`re correct, 2,6 bar pressure is what I have it set at, and that`s what made 0.3% difference on CO reading.
Potmeter is good (but also tried another one with same resistance area, to no avail as far as emissions go, 0.3% over-all is all it can affect)
I did have problems with the water temp sensor a year ago, it was broken, so I had cold idle hunting, soldered wires to the pins and no problems with that since

Potmeter is good, have not checked water temp sensor yet (Which I assume is OK since engine is not hunting at idle when motor is cold).

0.8 lambda and up to and over 8% CO is crazy rich, its overfueling like mad, without giving me grief while driving, idling or anything.
AND plugs which dont even have 1000km on them let alone 10000km, still look normal!
If it was exhaust leak I`d be getting over 1.0 Lambda value (but no better CO)

Have no spare MAP sensor but the one I have works (disconnect = engine stalls)

And yes I did visually check and re-grind earths for EFI on intake+head to firewall+block to frame-rail again (And measured them).