I have a feeling fitting SDO to Networkers could well be a difficult and expensive process, especially if going for something more than a very basic rudimentary system.
I haven't been following the SE lengthening upgrade in detail, so what are the remaining infrastructure obstacles? I know Charing Cross is a major issue as is Woolwich Dockyard. What else are we looking at, in particular that's difficult and/or expensive? It does seem to me that the ideal solution is simply to get all the infrastructure up to a point where SDO isn't necessary, but equally I'm aware that will have a price tag and disruption tag attached to it.
Platform lengthening is still the key one. On the Orpington to Victoria route, apart from Orpington, Petts Wood, Bromley South, Beckehnham J, the platforms are all 8 car with no room to extend. Same on the Catford Loop and Sevenoakes via Bat and Ball (although this is worked almost exclusively by Thameslink nowadays).
Victoria itself has a couple of 8 car platforms, due to short platform lengths and/or interference of long formations with entry/exit pointwork for adjacent platforms. Cannon St is all 12 car, Ch+ platforms 1,2, and 3 only are permissible for 12 car networker formations (with an awkward "alighting only" wooden extension on p3).
MOST other metro platforms are 12 car, with one or two exceptions, Woolwich Dockyard is a (ridiculously tight) 10 car and is boxed in by tunnels at either end. This is worked around by making 12 car networker diagrams skip it.
There have also been a few incidents with 12 stop car boards installed, unbelievably, at the wrong point, resulting in stop-shorts when the train was stopped at what was marked up as the correct point. Stock shortage is another issue.
In summary, it's all a bit of a mess, there are relatively few 12 car networker workings, restricted to the peaks, none on the Chatham mainline. It's difficult to see much expansion beyond this, far less an "all 12 car" metro railway for the foreseeable future without new (and a lot more) stock.