Code review comment for lp:~macslow/unity8/shellRotation

After cleaning tons of logging-noise from a succeeded and failed ap-test-run of ~/.cache/upstart/unity8.log the major difference I could find was...

succeeded run:
 * qtmir.sensor: Screen::customEvent - new orientation 8 handled

failed run:
 * qtmir.sensor: Screen::customEvent - new orientation 2 handled

So for the failing "right_up" (orientation: 8) case, it seems the orientation isn't touched at all or stuck at 2. Why/how that's happening I don't know yet.