From my experience, normally when nearby Richmond tracking for Bankstown, the usual need in normal circumstances is to get the aircraft down (i.e normal descent).
A normal basic descent point (without considering other factors) would be approx 30nm BK from 5000ft or 42nm from 7000 ft (5oofpm/180 kts). So, if you made the decision to return to Bankstown, for whatever reason, previous experiences would try to convince you that you need to start getting the aircraft down, and also that Bankstown should be easily reached with one engine out (and one functioning normally) if you are located on what would normally be a descent path.
I do note that if you were back at say 120 kts (or 107) with an engine out, the descent profile would be steeper at the same rate of descent 500fpm.
What I'm getting at is, under the pressure and workload of handling the failure, your gut instinct would tell you you need to start descending to make an efficient return to Bankstown. It seems in this case the decision to return to BK was made early, and a descent initiated fairly quickly. (Obviously, the most appropriate thing to do with an engine out is to keep as much height up your sleeve as possible and descend overhead, unless its on fire).
Note that if you are descending for most part, you won't have much indication of whether the aircraft can maintain height (closer to sea level)on one. (ATC asked whether the aircraft could maintain height but the reply couldn't clarify this as he was "on descent at this time"
Even though the instruction from ATC cleared to aircraft to 2500 (rather than instructed it to descend) it MAY have subtly reinforced in the pilot's mind that the aircraft should continue descent. (Note ATC tried to clarify as to whether the aircraft could maintain height just after this).
For whatever reason, the aircraft ended up at 1500 ft still at 12 miles out, requiring a LONG level segment, which the aircraft/pilot in the end could not achieve.
So to summarise my point, there would be a strong tendency to want to descend the aircraft, under the assumption the aircraft will hold level when needed) rather than maintain as much height up your sleeve as possible.
Our decisions are shaped, amongst other things, from past experience and knowledge of the environment we are working in. Having flown in and out of the area in question many times in PA31 aircraft (as I'm sure many of you have), my perceptions of the factors influencing the pilot's decision making will be different to those who have not, so I just wanted to add my contribution. People who haven't flown in this particular environment will have a more neutral/external perspective, but just as much to contribute to the discussion.