Only thing that stands out to me scanning the thread quickly is the relationship between knock "counts" and degrees retard. The 0-3 and 3-7 thing doesn't apply. It's strictly linear/proportional. There is a set number of degrees of timing pulled per "count." The system you mention was long thought to be the case, and since Todd Day (of TMO) put it out, no one argued. But now that there are many people manipulating the codes and burning their own chips (there is a whole yahoogroup for DSM ecu chip burners) they are finding that this really isn't the case on 1Gs either... Just a little history for you youngins
Nice thread, should answer a lot of questions for poeple!