@sunnyhighway: Without seeing the actual installation it's a bit of guesswork, but in general a bridge between panels is no problem. The ultrasonic sensors measure almost continuously, so a reasonably sized bridge will allow it to cross over. The angling around pipes has come up before, the friction is not really an issue, since the variation of length of the wire is within the robot (so the wire itself won't continuously slide up or down). What is an issue is the angle of the pipe to the solar panels, this will force the wire down and might cause snags. I guess some strategically placed guide rods would fix this, but this specific scenario might require some testing before stating en public that it's a done deal.
@Dave: I know the idea doesn't come to mind when buying the panels (believe me, I've spoken to many people telling me just that)!
About the strong winds: The Scrobby just fits under the ring, which has a triangular cross section, giving it some centring force (plus the ring mounts are at the outside, preventing it from sliding). When resting at the side of the panels the furthest extension wire has a very nice angle for taking up wind force towards the outside of the installation. Besides that I've done some consulting with an aerodynamics expert on this very subject (improving the body shape to prevent lifting and side displacement).
The wires extend from the body very close to each other, forming a hinge point. This in turn is used together with the compass and accelerometer to get an idea of current brush torque (the rotating of the body) vs expected and any bounces or wind gusts.