The AMMs on Motronics systems have to be the number one pattern failure out there. They always fail to under report (go lean) and the inaccuracy grows with the volume of airflow. In this case you are being compensated with an additional 22% fuel at the part load coditions the addaptation is done at. At full throttle it is probably considerably worse and thus not fully compensated, probably a cause of the misfire codes originally.

I would look again at your idle adaptation and verify whether it is plus or minus .5ms. The typical way an under reporting AMM gets adapted will have a large correction done multiplicatively to load (1.22 is large) and then at the much lower flow rate it will be more accurate and the universally applied 1.22 correction will leave it rich requiring a CTP (closed throttle position) correction subtracting fuel. If you in fact have a negative CTP value then the AMM is the likely candidate. It is less likely if you are straight line off.

Check archives for AMM and you will see that consistantly they are reading 1.25 and above in motronics cars that have been fixed with new units (of course they should be about 1.00 after repair)

__________________
Steve Brotherton
Continental Imports
Gainesville FL
Bosch Master, ASE Master, L1
33 years MB technician