What seems to be missed here is that for regional TOCs, rather than the London ones, passenger numbers never really dropped to the levels seen in and around London. Our services are vefting busier by the day and some are at pre-pandemic levels.
A service cut just to save money wouldn't work, as the cuts aren't needed here. So that doesn't magically save your day a week to make Sunday a part of it with the same number of crew.
Even if they are busy, they are unlikely to be paying their way unless they are 'standing room only' (and even then, only on lines with high p/mile fares). So as far as the DfT is concerned I think it may still be 'fair game' to cut back there.
1) peak services into the big cities, principally London. These trigger particularly high marginal operational costs (trains that might operate 2 services a day). The difficulty will be picking which services go. There is no doubt that the main suburban line into Waterloo will not need 18tph in future, but which 2-4 trains get removed?
Most of these have already been eliminated to be fair. But as you say they're very unlikely to return, regardless of what happens.
2) ‘competing’ services on the same route, often introduced as ORCATS raids. TPE north of Edinburgh being a good example. You might like to suggest your own candidates. My bugbear is the occasional SWR services to Bristol, or the GWR services to Brighton. Both cause a lot of extra cost for negligible benefit.
Didn't realise they ran north of Edinburgh
The ECML inherently lends itself to ORCATS raids under the current structure, but even under a unified 'banner' it doesn't seem unreasonable to want 2tph from Newcastle to each of London, Birmingham and Manchester (ideally clockface, obviously in practice it won't be).
In terms of withdrawing TPE north of Newcastle though, there will be little to no resultant saving for as long as the units are maintained at Craigentinny.
A more cost effective way of providing that may be to have some services splitting and joining at York, albeit that has its own foibles, and simply wouldn't work at the moment!
If there's any services ripe for removal on the grounds of duplication, it would surely be ECTL's, which will solely exist to compete with LNER and the airlines.
3) services in area services of maximum track capacity utilisation, where removing one or two services will have a disproportionately good effect on performance. Castlefield corridor is obvious, Birmingham - Wolves and the Cross City similarly, and countless places in the Home Counties.
The difficulty here is that you often end up with an unbalanced service if you remove a few to improve performance. E.g. the CrossCity is currently an ungainly 10-20-10-20 pattern because 2 out of a clockface 6tph have been removed. So a recast is needed to take full advantage of the benefits whilst providing an acceptable timetable.