At present, the .au namespace is divided into a small number of second level domains (2LDs). Those open to individuals and organisations are .asn.au and .org.au (for non-profits, etc), .com.au and .net.au (commercial entities and traders), and id.au (individuals).
This seems to be driven by the experience with new (and extremely expensive) top-level domains (gTLDs) that have been introduced over the last three years, along with the liberalisation of various country-code top level domains (ccTLDs) such as Singapore (.sg), the United Kingdom (.uk) and New Zealand (.nz).
Most obviously, the domain registration industry. The infrastructure is in place, and so an increase in the number of domains available for registration would presumably improve revenue and profits for very little new investment. According to the paper, "the majority of .uk direct registrations are held by the registrant of the matching co.uk name."
So if it were possible to register mcdonalds.au, it seems likely that it would end up with the well-known fast food chain, not to one of the businesses in the automotive, construction, electrical, health, plumbing, transport or other sectors that have 'McDonalds' in their name. Even if a relatively small company did manage to secure it initially, the temptation to part with it for a good payday would be considerable.
Furthermore, auDA's current policy would prevent the successful registrant of a new 2LD would not be allowed to 'subdivide' it by acting as a registrar outside the existing system. Some people will see that as an anti-competitive measure, others will see it as a way to ensure the technical stability and integrity of .au.
And if you see direct registrations in .au as a way of opening up the total domain space, note that the policy in New Zealand was that if an existing name in at least two second levels had gone to different registrants (a local equivalent would be withers.com.au and withers.net.au), then it is unavailable for registration in .nz (think withers.au) until those registrants have resolved the conflict between themselves - with no time limit. On top of that, in the absence of a conflict, established registrants in any of the pre-existing .nz 2LDs were given the first chance to register those names in .nz.
Back in Australia, according to the paper there has been little interest in the .melbourne and .sydney gTLDs, with less than 6,000 or 4,000 registrations respectively.
But the argument that this shows there will be little demand for defensive registrations isn't very convincing, especially when one of the possibilities is that registration requirements for .au will be more relaxed than for .com.au. Surely that would mean existing registrants couldn't rely on existing policies to protect their names, and so they would increasingly turn to defensive registrations in .au, even if they didn't feel the need to register the .net.au name that corresponds to their .com.au.
So apart from the off-chance of a windfall profit, it's hard to see how most organisations would benefit.
Would the proposal make life any easier or more convenient for the typical internet user? Probably not. It saves a few keystrokes - e.g., qantas.au vs qantas.com.au - but casual observation suggests that a lot of people just type qantas or another name into the box and let Google or Bing find the URL for them.
The answer to the question "who benefits?" (cui bono) might be found in another cliched expression: "follow the money."
If you'd like your opinion on the matter heard, read the issues paper [PDF] and then either make a written submission or complete the online survey (instructions are on the first page of the paper) by midnight on 1 June 2015.