Ugh, I can't believe you actually swallowed that piece of Liberal Party propaganda. The 90 billion estimate that the Liberals are throwing around requires ALL FOUR of the following assumptions to be true:
1. Revenue (ARPU) growth is to be significantly lower than NBNCo projections (for their estimates, they assume that ARPU growth is only 3.5% per annum).
2. There has to be a minimum 40% cost blow-out in the average per-premise cost of FTTP installation across the duration of the project.
3. There has to be insufficient take-up of fixed line services (i.e. a significant portion of households have to go "all wireless").
4. The project is completed by 2025 rather than 2021.
While I admit that number 4 is a remote possibility (NBNCo is already 3 months behind schedule), number 3 is ludicrous (a significant portion of wireless users in Australia are on wireless simply because their fixed line service is useless), there is currently no evidence to suggest that blowouts on such a scale will occur, and statistics on NBN seem to suggest that NBNCo may have underestimated ARPU growth.
Also, the NBN isn't being paid for with taxes - it's being paid for with government bonds. Considering that Australia has a AAA credit rating, borrowing for an essential public infrastructure project that will create a profitable national asset is hardly irresponsible - in fact it's a sound investment. On the other hand, assuming that our existing copper network can still be used as meet our future needs is highly irresponsible. Just remember this - in 2003, Telstra told a Senate inquiry that the copper network would need replacing in 15 years. The Coalition wants to keep using this same copper well past 2019.