It takes everything into account except contamination (bad air filter) and maybe a couple of other factors that it can't model with the sensors (coolant leaks/head gasket/intake gasket). It obviously does a much better job than a static "2 sizes fit all" and the numb safety factors that they have to include to make them valid. The current recommendations are some mean average of some polarized extremes.
Now sure, it may be hard for anyone to do over 10k with an OLM with conventional oil ..but you can put a synthetic in if it gives you confidence ...or, as I said, just do 80% or whatever for you "giving my car the best protection" quotient.
Quote:
oil life is not linear.
I'd say that oil life would appear linear to a certain point. The oil will have a reasonable limit on suspending particles before it starts depositing them. TBN would probably decay on some "half the distance to the goal line" manner after the buffering is gone..but it's hard to say since most of us span many seasons in any OCI that would tax TBN to the point of it being an issue. The process variable is variable.
As far as the rhetoric goes, the only linear decay that's used by the OLM is zddp depletion. The other factors are accumulative in decrementing the indication.
