Re: Pitfalls of G.fast
It's both. Emergency Power & Maintenance.
I don't see BT been allowed to self power devices from subscribers homes/offices as these would sit outside an area that get regular inspection/outside an area of BT's control.
There is potential for that device to send 240VAC onto the BT copper pair. i.e. You'd have to have written permission from the subscriber for regular yearly access to check the Power Supply (fed via unverifed consumer 240VAC) is working correctly, there would need to be strict conditions. There will substantial costs involved.
Outside an area of Control.
BT can't control the conditions that device operates in, temperature, damp, vacant landlords/owners etc, yet it feeds power (originally from an unverified Consumer 240VAC source) into BT's network. I doubt it will happen.
If it does, it negates much of the cost savings of having Passive Fibre optic (FTTP) all the way to the subscriber v so called cost savings of Powered devices/G.fast. BT's only way is to 'BT self power' these devices through its own equipment within its network, I can't even see them been allowed to use 2 pair copper to do this from their own PSUs. The reality been, most G.fast nodes will need to be connected individually to the grid.
As said, G.fast gets exponentially expensive, depending on the blanket " 'upto' coverage" you are looking to offer.
Of course, G.fast isn't designed to be about blanket coverage, its about selective coverage - this is BT.
BT want a restrictive tap (as in water) between the subscriber and BT, to gouge its customers, by restricting the best throughput speeds through selective pricing. Its more difficult to do that if each subscriber has a passive fibre optic running into their house, why's this resource limited? where's the restriction exactly?
G.fast is designed to be that artificial restriction, to make Ultrafast Fibre seems like a 'limited resource', yet the actual unrestricted fibre sits 20 metres from the subscriber on a Pole, with a active G.fast 'tap' device sat between it and the subscriber.
Much of BT's argument can be shown to be a lie, 'Passive Fibre is expensive, Actively Powered Copper Piggyback Technologies like G.fast are cheap', its been conditioned into ofcom/Politician's little heads i.e Ed Vaizey. This is even before you factor in costs from Interferene/Cross talk/Low Frequency Pump/Induction Noise + fault finding costs.
When you factor in the true costs, long term maintenance, it can be seen that BT's are biasing the technical reasoning towards technology which favours BT's legacy copper carcass, and only BT. BT are the only ones not seeing their Copper Network only in terms of its scrap value, post Brexit. Mediocre won't do in this new era.
The local loop / BTOpenreach needs to be completely separated and that probably means taken back into public ownership.