The NBN debate is a mostly tribal ideological crap. Better than a deep analysis would be to knock some of the dumb ideas out of the "debate"
1. Can wireless ever replace a cable/optic infrastructure? No.
Data throughput is of the order of available bandwidth. This is textbook physics. Get over it. There are loads of noisy dills "participating" in the debate who don't understand this. Techno golden age mythology won't trump information physics. If we could just get these wireless nuts to shut up the debate could move on to real issues.
There are strategies for dealing with radio bandwidth limits. Build more base stations. They need to have low power to share the bandwidth. Low power signals don't go far and can't go through walls. That's why it is smart to place them in our houses and supply them with high bandwidth cable links. Another strategy is directional antennas. Reuse bandwidth in different directions. Designs for these systems are around but they are big and expensive. There is a trade off between cost/size and directional resolution. You can get phenomenal radio resolution out of the Square Kilometre Array but it costs billions.
The other strategy is to use cables to channel the signal. This works well.
2. How much data do we need? Funnily enough, this question is hardly discussed probably because it's speculative - and because people are innumerate anyway - but it's important. The ways you might guess this are to look at the household bandwidth usage, currently growing exponentially, and project out say 20 years. Presumably it will taper off it is currently showing no signs. Any realist guess based on this chartist's method says fibre. You could also consider applications. Video is the big hog. There are limits to detectable resolution and useful screen size but if you take a few screens of live video at twice the current screen size at bluray resolution and faster frame rates you've blown ADSL (or your own personal phone tower) and need a better cable. Can it work on a fast copper cable system? Maybe, but you're up against the cable limits. Fibre will handle anything we try to throw through it. If something like high-res 3D video becomes the norm the chances are you'll really need the fibre.
3. Economics I. This is a big public infrastructure project, like electricity, reticulated water, roads, the sewerage system, the phone system. The cost of digging up the roads and laying new cables is significant, like a couple of thousand dollars per premises, depending on where, how long, etc. Reality check: This is like 10% of the cost of a low-end new car or 25% of the cost of a cheap kitchen renovation or less than 1% of your typical home price. People spend money like this on a new TV that will be chucked in maybe 5 years. NBN fibre will last 50 years plus. (The routers and transceivers won't, but they are cheap.) The headline project figures are tens of billions but per house it's really not that much for something you use all the time. Worth the money? Well, that depends, but while it is big but it's not a major life purchase.
4. Economics II. Can the private sector do it? Yes. Can they do it for nothing? No. It's about the same cost either way but if the cable is privately owned they won't itemise the cable cost in your bills. If a company owns the only cable to your house, would they like to screw you forever? Yes, that's how the market system works.
Alternately, insert your own ideologically motivated answer from the government-is-wrong or capitalism-is-wrong myths.
5. The fast copper v fibre trade off.
Initial Cost - Cheaper by 30 to 40% upfront
Maintenance cost - High
Speed- Good enough for current applications, probably too slow
Lifetime- 10 to 20 years. We hit bandwidth limits and/or maintenance cost before too long
Initial Cost - Higher by 30 to 40% upfront
Maintenance cost - Low
Speed- Fast enough for any foreseeable applications
Lifetime- at least 50 years
Take your pick.