Originally Posted By: bulwnkl
KLowD9x, if you could truly measure a shift quality improvement, then you might want to re-think the 'sensibility' of your 75k mile replacements. I agree with unDummy here; you shouldn't really be able to tell the difference before vs. after. If you can, you probably already waited too long. This assumes it's a genuine difference rather than a mind game you're playing with yourself, of course.
As to your original question: If you will only use 1 criterion to the exclusion of all else, and if it must be only 1 of the 3 you proposed, then mileage is the best within those parameters.
The only time I could measure a difference in shift quality was after the first flush. I flushed through 16 liters of fluid at 50K miles because the fluid was black. Which is very common for this particular transaxle at that mileage. However, after I performed the fluid exchange, the 1-2 and 2-3 shifts were quite stiff. Not hard. I could feel them engage, which is odd for this AW box.
The only reason it made a difference this last time is because I reset the adaptive data and performed a relearn test drive. I'm sure if I had done this same procedure when I performed the first flush, I never would have noticed a difference in shift quality.
All I know is that the fluid that came out after 70K miles was in good condition, smelled new, was still red, and was just a little cloudy. 50K miles seems like a reasonable time to do a small 8 liter flush from now on.
The best part is that, just like when the car was new, I can no longer feel the shifts when the fluid is hot. (When cold, there is a different shift program to make the engine warm up faster. It's annoying, but I guess they had to do it for the sake of emissions.)