You've hit the nail on the head Mike. Basically the 2G MAS has a lower corresponding Hz reading, so when you put it into a 1G, the ECU thinks there's less air going into the car than actual, so it'll lean things out. Then the 550s richen things up and it balances out.
I've posted the question on the DSMlink forums, trying to see what the actual airflow cutoff point is. I mean, if you can't trust people who've actually read and modify the ECU code, the only alternative is to download the code off the chip directly...
Basically, from my understanding, fuel cut was implemented to prevent airflow in excess of what the mitsu engineers felt the fuel system could reliably support. So, the only way around it is to lower the perceived airflow and richen it up via bigger injectors, ala SAFC, MAF hacking, or take the limit directly out of the ECU, ala TMO, DSMlink.
If you hack your MAF, make sure you have a way to compensate, i.e. higher fuel pressure or bigger injectors. BUT, then you are looking at a non-linear flow bypass, i.e. you'll get different proportions of the flow by-passing the measurement volume at different flow regimes, which makes it hard to tune right.