Originally Posted by JCL

No flaming. However, if we believed it was the fluid, then the high mileage examples without failures would tend to disprove the theory. Fluid failures don't tend to be random in my experience, they are consequential rather than being the source problem.

A failure related to fluid (whether the fluid caused it, or was damaged by slipping clutches and overheating) involves burnt fluid, and signs of overheating. We don't hear about transmissions being worn out, rather transmissions breaking and not shifting properly. The damage from the ones we have seen is limited to a single clutch pack or actuator, and not pervasive throughout the transmission. A failure caused by fluid would tend to affect all clutch packs. There have been some examples of the ZF transmission having a specific failure of one part (swissfrank has the best post on this) That could be a manufacturing defect, or a design defect, but the failures seem to have additional modes as well other than this specific point.

Driving conditions is an obvious explanation, but the number of examples of pulling large trailers for extended distances tends to negate overloading/overheating as a theory. Extensive manual shifting could be a factor, hard to say.

I come back to the electronics/control package. BMW puts their own controls in these transmissions, to a certain extent. I don't know where the line is drawn between BMW supply and GM/ZF supply, but there is a line. This fits with why the GM and ZF transmissions have similar failure rates. Think about it: what fails randomly on a BMW? Software, sensors, etc, are the most likely cause. When one of those items fails, the affected system stops working, whether it is a clutch pack controlled by a pressure sensor, or a speed sensor, or what have you. Just my theory.