Maintaining a token diesel fleet costs money, and the political benefits from electrification are huge.
If a rolling programme is such a boon as people around here say it is (often to justify not accelerating the programme) then there is no reason not to continue until the job is done on both NR and NIR metals.
Also, there are only a handful of lines with only 'a couple of trains per day' and if it wasn't for stock limitations and suchlike there would be a case for many more trains on most of them.
Additionally the elimination of traditional signal boxes in favour of integrated signalling centres will enable lines that currently cannot justify staffing the boxes to remain open since the marginal staff costs are drastically lower.
There is no economically optimal solution to a high speed railway. All costs must be considered, environmental included.
Environmental costs are not the same as physical costs for the reasons discussed here.
We have to make a 'value judgement' and it is possible to generate defensible figures that dissagree by an order of magnitude or more.
There is no technical reason why you could not traverse the Chilterns on a modern high speed line without a single tunnel thanks to the very low cost of earthworks with modern earthmoving technology and the fact that HSR trainsets have been demonstrated in service on 4% gradients (in Germany).
Therefore the cost of all tunnels in the Chiltern area must be ascribed to the NIMBYs/environmental costs.