I knew from the manual that locking action under a leader fall is only certain for single ropes or both strands of a double, but I did think at first that would lock off on rappels, until I thought about how the mechanism worked and conducted a little experiment with low rope weight.
Well, it's tricky, because it *kind of* auto-locks. Kind of is not what most people want, although if you understand the limitations, I think it's actually a nice balance between mutually exclusive things (dynamic catch and auto-locking).
My understanding is that the grooving problem has been addressed by the addition of brass bearings on the part of the cross piece that contacts the slotted plate.
PMI, who distributes them in the US, was nice enough to sell me one of the new ones at a discount. It *looks* like the brass sliders may fix the problem I had, but I'll let you know in a year or two.