And so, after almost a week (!) of investigation, and a very fine Cruze loaner, the dealer concluded there was nothing abnormal about the OLM. I guess I'm not altogether surprised given how the oil looked, but it seems that my lack of any highway driving whatsoever since the oil change is what was doing me in. Dealer's body shop also fixed a minor scuff on the front bumper they put there the last service (and you can't tell where it was), and finally seems to have gotten to the bottom of an annoying dashboard rattle I've had intermittently since Day 1.
One thing I didn't ask and probably should have is whether the lack of highway driving has a cumulative effect on the OLM algorithm/regression coefficient (whatever math they use). The run down of the percentage seemed to be accelerating, until the highway drive to the dealer.
Incidentally I have to say that after a week with the Cruze, I'd probably get that car knowing what I now know. But this time a year ago, I really didn't want to deal with anything bigger than what I already had. The Cruze (1LT) is a bit underpowered, but a phenomenally comfortable car, especially WRT the seats, pedal placement, ride and quietness, and your attention is only called to the underpowering when going up a significant hill; for regular stuff like pulling away quickly from lights, it's very good. But I'd miss the Sonic's visibility, crash avoidance system and double glovebox.
