It is actually quite clever, it is 25Kv, but it is 180 degrees out of phase with the contact wire. This effectively makes it a 50Kv system, with a resultant increase in power available, and/or increase in the distance apart feeder stations can be, which means you need fewer of them. You still only need 25Kv clearance to earth and 25kv insulators, but you must have 50Kv clearance from this cable to any other live part of the OLE. However getting 25Kv clearance to earth under bridges is a bit more tricky compared to the tensioned OLE, so they tend to be run as heavily insulated cables in troughing under bridges. They also in theory perform the same function as the return conductor in reducing inductance/interferance in the signalling cables. However modern traction and signalling systems tend to be much better in this respect compared to 1960's technology, so it is likely that existing return conductor systems may be removed as they will no longer be necessary.
One other innovation that will become common place in the future is invertor feeder stations. It has always been a bit awkward in that the railway imposes a heavy single phase demand on a three phase national grid. They try to mitigate this by having adjacent feeder stations on different phases, but it is far from ideal. With an invertor feeder station you can take load from each phase evenly, rectify it to dc and then invert it to single phase AC. Not only does this eliminate asymmetric loading on the national grid, but all feeder stations can output on the same phase as each other eliminating the need for neutral sections, since the output phasing is independent of the input. I think they also make regenerative braking feeding back in to the grid easier as well.