Well what we do know is that openreach are prepared to extend fibre to street distribution points without any guarantee of orders, they already consider that viable.
That is not what BT or Openreach have actually said. They have been very careful to say any rollout is 'subject to xxxx small print' and neither have they actually said that G.Fast will go to the distribution point-in fact most now expect it to be offered from existing FTTC cabinet locations first.
I'd make this same point: BT can likely reach their 2020 target of 10m homes without taking fibre any deeper at all; just by placing a G.fast node alongside the FTTC cabinet.
And even when they want to go deeper, the Huntingdon and Gosforth trials are meant to help determine where (between PCP and DP) they will choose to site the DPU's.
BT haven't committed to where they will plant the DPU's yet - but they have said that putting 4 million of them out there (which is what would be needed to site them at every DP) is too expensive, so we can expect them closer to the PCP. Their work at "persuading" chipset firms work on extended range options for G.fast suggest they may be looking at copper distances of over 200m - perhaps 300m or 400m.
If fibre is pushed out that far, the extra cost of FTTP on top of that is not that significant, All openreach would need to do is push out fibre the same amount as is supposedly planned for g.fast and then only do the rest when an order comes in, so there is no rolling out FTTP without orders, in effect similar to the FTTPoD, but the main difference been the product is actually marketed and priced at a point takeup will occur.
If fibre is pushed out to the DP, then yes, rolling out FTTP is cheaper than it would otherwise be but you are wrong in thinking the costs are not that significant. The costs of converting homes to fibre are more than connecting a DP to a fibre node - and doing it on demand also more than doing an entire area at once.
Again, I'd make roughly the same point. However, even if BT chose to site the DPU at the DP, it would be a case of deploying 4 million DPU's, along with an average of 350m of fibre for each one, running the fibre mostly through ducted access. That's a maximum of 1.4 million km - but likely to be a lot less, as routes from PCP to DP can be shared/daisy-chained.
Rolling out FTTP would need to be to 28 million premises - a large factor more than 4 million - and an extra 50m to each: which turns out to be another 1.4 million km - but this time unlikely to be shared. All with the extra admin of needing appointments.
IIRC, the calculations done in the study on Amsterdam suggested that deployment of G.fast would be less than half that of FTTP.
And, IIRC again, the Nesta calculation of deploying G.fast vs FTTP, and factoring the eventual reuse of deeper fibre, was that deploying G.fast *now* followed by FTTP in 2023 cost the same as deploying FTTP now. Their deciding factor thus becomes: If G.fast gives you all the speeds you need until 2023, or later, you might as well deploy G.fast. If it needs replacing before 2023, then you should go for FTTP.
Meanwhile, the BSG base calculations (IIRC, again) suggested that average home would be fine with 19Mbps in 2023 (ie 19Mbps isn't enough for 50%); that 38Mbps was needed to keep the top 1% of homes happy apart from the busiest minute, and that 50Mbps was needed to keep that top 1% happy at all times.
BT themselves were estimating that, in 2025, "Many homes, most days, and some of the time" would be using 50Mbps.
If BT believe any of those figures even vaguely, G.fast looks economically worthwhile. Especially if they can pass on the costs of FTTP to subscribers who demand more than a G.fast speed, and will pay a semi-reasonable on-demand charge.
Also back to the topic at hand, I wonder how many of EE customers (now owned by BT) will move to BT retail broadband/tv products when they get bundling offers thrown at them.
But I don't think the change will be as great as you expect given that EE already offer EE TV, line rental and broadband to their customers. The main opportunities come from BT being able to offer better mobile contracts to its own customer base.
BT already do a quasi-bundle for their mobile offers, by knocking £5 off each SIM per month, if you have a BT broadband contract. But I agree that the short-term opportunity comes in improving that deal.
I also wonder whether they'll follow DT's lead, and market a home broadband service that can combine 4G and "fibre" connectivity for backhaul, giving faster speeds than available purely through the fixed network. It might even allow them to reach more of their USC target properties at 2Mbps, or the USO target of 10Mbps ... possibly in combination with the rollout of the emergency services hardware.
But I think the ultimate target is in a fully converged network. Ubiquitous coverage. Where your device (and your family's devices) connect over a BT connection, whether that is your home fixed-line connection, a work fixed-line connection, a BT WiFi connection (via someone else's home/work connection), or their 4G LTE network. Or even a combination of these simultaneously.