Uh uh. Your contention is that UoM is 'doing something right' based on those stats? Whilst I don't disagree with the Melbourne Model on principle (I just wished it was offered as an option rather than a requirement), and it would become really really good once it is fixed and refined a little to reflect the American system more, I don't think those stats prove that Melb is doing something right.
ATAR cutoffs depend on three things: supply, demand and minimum ATAR set by uni - not necessarily the quality of the course. As supply increases and more students are able to get in, ATAR cutoff would naturally decrease, irrespective of whether the course is good or not. As demand increases (which may be due to the quality of course, reputation of the uni or even just clever marketing strategies, or just the fact that Melbourne is at an overall very convenient location for people all over the metro area - this almost convinced me to choose Melbourne BSc+MD over Monash MBBS) the ATAR cutoff would naturally increase. Minimum ATAR set by uni may or may not reflect course quality/demand. It may set a high ATAR minimum because the course is genuinely difficult, or simply as a marketing tool (excellent example is Biomedicine, which is essentially an integrated science degree where most of ur subjects are picked for you, and has a minimum ATAR of 95, as opposed to Science's 85).
So Argo, Melbourne may be doing something right, it may be doing something wrong, but your ATAR stats don't prove your case.