FlightPath,
It looks weird because of the perspective of the photo. Here is another photo taken of the same airplane.
http://i.imgur.com/uuqth6O.jpg
You can see the right winglet is nicely tilted upwards and not "curled-over" like your photo makes it look like. The left winglet explains the discrepancy between your photo and mine - the winglet is actually curved backwards towards the tail. It seems logical that this would help reduce induced drag as the winglet is affecting the air for a longer time, reducing the intensity of the wingtip vortex. It'd be interested in some aerodynamic analysis of this but I doubt Airbus will be revealing any of that data!