This ties in with the 2005 comment by Sol Truijillo that since 2000, Telstra had under-invested by $2-3 billion in Capital and Operational works (maintenance was let go). The real situtation would be worse.
NBN Co has leveraged its GIS database and I.T. by using Optimisation software for network route planning.
It has shortened the planning time from 145 to 20 days according to NBN Co. Though the vendor claims much better speed-ups.
Importantly, with the more open constraints of Fibre, the software can find better fine-detail solutions than people, leaving the network designers able to try multiple strategies and designs to better fit individual areas and local conditions.
Replanning areas, even in progress, using new rules or to accommodate changes and new developments is a much simpler task.
In their 2007 submission to the ACCC on the G9/Terria "FANOC" network proposal (ADSL2), Telstra commented they could optimise the copper network for an FTTN and reduce the number of nodes by 40% compared to the "leave the copper as it was for phones" approach by G9/Terria.
I've no idea of what benefits optimisation of network layout for Fibre will yield NBN Co, but suspect it is large and worthwhile doing on many levels, not just reducing the kilometers of cable needed.