The composite photo included in the main paper by White and colleagues (2009) is very poorly scaled. For example, the measurements given by Lovejoy and colleagues (2009) make the radius 95 percent the length of the femur, but the photo renders the radius quite a bit longer than the tibia:
This has the effect of making the photo look much more like the proportions of a chimpanzee or gorilla (or Oreopithecus...) than is the case from the skeletal measurements.
Bottom line: don't use the photo in classes to talk about limb proportions; use the Matternes reconstructions which appear to be correctly scaled.
UPDATE (2009-10-04): Although I will say that the Matternes reconstruction has much more slender-looking hindlimb bones than the photos. Since one of the interesting aspects of Ardi is the large reconstructed mass (51 kg), I find the very gracile long bone reconstruction misleading, also.
