A rational explanation
1. That Telstra would have the “option to compete with the NBN” while receiving funds from it.
The NBNCo Corporate Plan (2013) states that in April 2013, 47% of fibre connections were 12Mbps. The fact it costs only $5 extra a month for 25Mbps, suggests to me that the 47% either don't care about their network or want the cheapest connection possible. If we assume that the vast majority of these people have a smartphone or will within the next five years, then I would argue that a reasonable number of these customers could be poached by the mobile network operators on the basis of price, especially when for a reaonable number 4G will be faster.
2. That “projected customer take-up and average return per user were likely to be proven optimistic in such a competitive industry”.
NBNCo is based on the assumption that they can drive ARPU from just over $20 to well over $100/month as demand rises. This may prove challenging. The high connection charges (AVC) will present a barrier.
3. “Lazard … pointed to anecdotal evidence that 20 per cent of premises could ultimately be wireless-only.”
The NBNCo Corporate Plan estimates 16% of premises will be wireless only because it is cheaper, so 20% is not unrealistic. NBNCo are estimating that only 70% of premises will connect for a variety of reasons.
If you are considering replying that mobile wireless is not adequate for internet, please note that I agree with you, but then I would select the 100Mbps plan not the 12Mbps plan. To understand if wireless is adequate or not you first need to explain why 47% would select 12Mbps when for $5 extra they could have 25Mbps.