Having read the Economist Intelligence Unit's brief accessible summary on broadband plans compared among dozens of countries, it has become obvious to me. Australia's broadband plans have been scored low because of the high comparative cost and mediocre quoted speed.
The issue I have (as I had with their take on university privatisation) is how the length of term of investment is never even considered as an issue. Wireless (Next G as oposed to Wi-Fi) has excellent returns on investment in the short term, but is in no way future-proof.
The real crux of the issue - Who should pay for and/or own infrastructure as opposed to running costs and retail sale of bandwidth etc, is never mentioned. It wouldn't be an issue to me, if the alternative suggestion was the government owning and spending money on a much reduced infrastructure - That would be a reasonable alternative, but in the long run, it wouldn't actually be cheaper. To me it is like Mr Windsor said - "You do it once, and you do it with fibre".
The biggest returns will happen once Moore's law catches up with the capacity of the network. Memory capacity still doubles every 18 months, but data generation and traffic on the internet is doubling quicker than that, about every 12 months. Therefore, given that if you build it, the traffic will eventually fill it - it makes sense to optimise the cost of the end-point infrastructure, rather than requiring a shorter timeframe to gain a payoff with something that will need to be upgraded again, duplicating a lot of the initial work.