If people don't use it!
Just because it can be built doesn't mean there is a market for it.
It only takes another technology that is 80% as fast 100 time more mobile and 1/2 the price and its redundant.
CD versus MP3 is a classic example. MP3 is a truly rotten system in a purely technical sense in terms of the quality of music reproduced, but it has so many other advantages that it has been widely and rapidly adopted.
I'd even argue that CD's really only replaced vinyl not due to sound quality, but due to other factors. No more messing about cleaning records, no more hunting for the right stylus when it wears out, no more jumps and it's a lot smaller and easier to use (especially when you don't want to play the entire album from start to finish). Same with cassettes and all the hassles they came with.
For most consumers, that sound quality is better was simply a bonus of switching to CD - they don't have good enough speakers to get the full sound quality anyway. I'd argue it was convenience and marketing, not audio quality, that saw the mass adoption of CD's.
Give consumers broadband that is
mobile at the same price but with 50% lower speed and it will kill wired connections for most users just as CD killed vinyl for all but specialised users. Convenience, marketing etc tends to win out over pure technical performance anyday given that most users won't push the technical limits anyway..
It was the same with Beta (technically superior in just about every way) versus VHS in the 1980's. Same with just about everything - the mass market doesn't automatically adopt the technically superior method unless it also happens to be the most convenient and best marketed.
All that said, some users do have a need for robust communications. A classic example would be that the that the physical part of the NBN that actually exists to date and is in operation was built by and for the use of the electricity industry. That's up and running in Tasmania and has been for quite a while - a need existed and so it was built.
Letting others use it as an alternative to Telstra wasn't hard, if your're going to have a cable then making it bigger doesn't add much to the cost, but getting the "last mile" to the household was always the problem financially. Hence the (unsuccessful) attempts at using existing powerlines for that purpose.
And the NBN will have the same financial problem. Bulk data between major points is the profitable and easy bit, getting it to individual small consumers is the expensive bit. It's comparable to electricity, transport, gas or anything else - the rate you pay at home is several times higher than the delivered cost (including production, transmission / trasport etc) to the sub-station or warehouse that's just a short walk from your door. It's the last mile that costs, and that's because there's no real volume there - lack of scale economies.
As for wireless, my own experience (at work) with it over a large number of geographically dispersed sites is that it's useful but never, ever rely on it for anything that actually matters. Good enough to send emails to your friends but usless at present for controlling critical infrastructure unless it's some situation where communication doesn't need to actually be reliable.
As for the NBN, we'll still need all but the "last mile to the home" of cable for wireless technology so it won't all be wasted. But I can't see it ever being finanically competitive as a means of connecting individual households - "the last mile".