It's all a complete mess on the eastern side of the Pennines. The existing infrastructure follows the geography, but that geography really doesn't help the idea of economic integration between the regions. Going against geography is clearly possible, but it'll come at a fairly high cost. That cost will probably be too high for even the loosened BCR calculations the politicians seem keen on for the NPR project.
Part of the mistake seems to be this idea that NPR needs to be a new and separate rail service, like HS2. I don't think that is realistic for regional services in the north. NPR really just needs to be a massively enhanced TransPennine Express network. It can't be expected to be all things to all people, as any new infrastructure would be better targeted at doing one thing well rather than a lot of things poorly. The sort of infrastructure you want to run a fast service to London is not the same as the sort of infrastructure you want to run a cross-Pennine or inter-regional service within the North. With the spread of population and economic activity across the region (even if Leeds and Sheffield city centre might be the strongest, there will still be massive demand for park and ride railheads) I think the end solution will balance more towards the NPR idea than HS2.
Probably the most transformational thing NPR could concentrate on is fast entrances and exits from city centre stations. 200m classic-compatible units seem to be a fairly reasonable model to serve most urban regions for their London links. None of the individual places east of the Pennines is likely going to be able to fill a 400m unit, unlike Manchester (+Airport) or Birmingham (+Airport). Focussing on these capacity-constrained sections may well mean it's possible to make use of existing routes which serve existing railheads. It's entirely possible to build small sections of bypass track specced for 100-140mph, joining and leaving the existing lines where appropriate.
Looking at a map now, I'm thinking whether it's best to send Phase 2b to Doncaster and then use the fairly high quality (and for some, only recently built) section of the ECML between there and York. If more capacity is required on this section, that could be done by adding another pair of tracks alongside. The alignment could support higher speeds on new tracks with modern structures, OHLE and in-cab signalling, and is pretty close to the route as the crow flies. That could be built as a separate project if and when the time comes, without requiring a fundamental restructuring of the route network.
The money saved would be used instead to upgrade the route north of Chesterfield, through Sheffield and across to Doncaster. That would mean a good quality express route all the way from York through the core urban areas and interchanges of South Yorkshire. This would be available and useful for 100mph express regional services as well as specific NPR and HS2 classic-compatible services. A similar set of upgrades, probably involving reasonable sections of new line bypasses, would link this Sheffield-Doncaster axis to Leeds via the existing Wakefield railhead (HS2 services to Leeds would run straight onto this rather than going via Sheffield or Doncaster). Probably most important would be a new joint approach to Leeds for express trains from Wakefield and Huddersfield, allowing the existing route to be metro-ified.
I think it would be really quite hard to manage a programme like that as a single project like HS2. It would mean a series of individual but fairly major interventions. The project split point with HS2 would probably have to be the Chesterfield and Doncaster connections to the NR network plus a spur to be extended towards West Yorkshire. After that, it'd be up to whatever transport body is best placed to design a new network for the north of England. It'd be a mess compared to the original, simple plans for Phase 2 but it'd be much more reflective of the real complexity of the area's transport needs.
