In the last thread on the subject I was only using the AFM for adjusting the fuel mixture. It looks like it also affects ignition timing. All of my efforts to keep it from opening up all the way seemed to trick the ECU into using a partial load portion of the ignition maps, where it was actually running under full load. The chip I'm using is set for "10 degrees btdc" per the seller, but only at full load.
So I ended up loosening the AFM, making an even smaller signal line restrictor by squeezing a vacuum hose connector with hot needle nose pliers, opening the clippard valve almost all the way and backing off the center screw to just a little preload. In theory it should be opening up the AFM sooner and further than it was before. This kicks in more fuel and retards the ignition timing for full load operation. I took the extra fuel out with the in line restrictor and other adjustments, and its much happier with 10+ psi for it.
If that theory is correct, fitting a larger AFM that wouldn't open as far would have the ECU running on the partial load area of the ignition and fuel maps, despite the TPS signal. It would be closer on the fuel mixture, but too aggressive on the ignition timing. Same for a MAF conversion. It would be great for tuning fuel, but theres no way to force the ecu to run on the full load side of the ignition tables while using the partial load fuel tables. It uses the AFM signal to determine both outputs. So at least in my case, the AFM was more important for ignition timing than it was for fuel mixture adjustments. There are other ways to adjust the mixture, but the ignition timing is very precisely controlled. Once I went over 9 psi, the AFM needed to be set for ignition timing, not fuel.
And as far as I know, still no limp modes.