Does the ECU reduce timing in a specified number of degrees per retard event or is it a single, continuous, linear retard until the knock sensor goes dark?

My car experience with a Bosch ECU is 3 degree increments / count knocks / 3 degrees / count / etc until the knock count is < 100/sec.

 

If ambient air tempos contributing to spark knock an oil cooler isn't going to help. You would need an intercooler to cool the incoming air charge. That being said an oil cooler is money well spent. I am buying a jagg 7 row and putting a manual bypass switch on instead of a thermostat.

 

It will just pull more than needed (up to the max specified if your tuning device has this table). Then it will decay back to zero if there is no additional knock. Another table that comes into play is the adaptive spark control. The delphi ecm will learn a value to retard the timing. These values are retained between key cycles and will slowly return to "0" as long as there is no more activity. Again not all tuning devices have access to these tables. If you do have access, this table should be set to zero while "working" the timing tables, as they will influence your results. If you get the TTS tuning guide it has great explanations on all the different tables.

 

I've been studying afr and timing and find this thread very helpful.
Looking at various maps that tmax has on-line it appears their arf run somewhat rich. Living in NorCal our fuel can be questionable. Leaning poor quality fuel can cause all sorts of problems. I'm thinking oil temp gauge and oil cooler is necessary for anyone desiring adjusting afr to leaner condition, especially in summer months.
I also think ecm's such as tmax are great in allowing owners to load different maps for seasonal, or type of riding (1 or 2 up, hwy cruising, or twisties). A guy can have several maps to load, initialize run iac, and go.
I have no stake in any company, just a retired engineer that likes to optimize my vehicles.