The question is whether running a long, quite slow service is the best use of any resources that are provided to the railway.
I'd argue that "metro-isation" has proven enormously succesful repeatedly over the years. Simple timetables operated at the highest possible intensity have been shown to grow traffic over and over again.
Indeed the current overcrowding on XC is largely a result of this suceeding with the Operation Princess concept, even if it was cut back.
Indeed, I do wonder if there's not a case for a rolling trail of running +1tph on route with low frequency to see if it generates enough extra passengers to justify it being kept.
For example, start with a list of routes where there's 1tph or 2tph and there's capacity to run more services (even if that's only to a large town to change to get to the major city) and starting with places with reasonable train loading trail for 18 months +1tph (or even +0.5tph).
The most likely to succeed are those which would run across a typical interchange station.
If it shows a decent level of passenger uplift (i.e. enough, or at least nearly enough, to justify the extra cost of running those services) then keep it, if not review to see why and either assist the trail or revert back to the old timetable.
Once you start getting to the point where the list is looking less likely to create services which would result in the services being maintained, then look at services with 3tph and 4tph and so the same again.
You might end up with some "odd" start/end points due to capacity issues.
For example (obviously subject to capacity) you could have a service running Andover to Farnborough or Basingstoke to Bracknell (as both would link services either side of a interchange station - Basingstoke and Reading).
As Andover to (nearly) Basingstoke is 2tph in each direction, so would be easy to path additional services (the tricky bit is the turning around), Basingstoke might be an issue (mostly heading westbound due to platform 1 being occupied far more than platform 4) but then the show lines up until Farnborough aren't that busy. Again finding a time to turn 1tph at Farnborough might be a little tricky.
Likewise, Basingstoke (platform 5) to Reading is likely to be fine other than the junction South of Reading (in time there might be a car for a platform 6 at Basingstoke, but that might not be needed for the trail). Through Reading might be tricky, as could through Wokingham. However, through to Bracknell should be fine, until you then look for a gap to reverse the service.
As it looks like there's no additional infrastructure required to turn the trains and so as long as the cost to run the trains is covered, the extra wear from the extra trains would likely be covered by the infrastructure charges (unlike a new line where there would be additional fixed costs in addition to the variable costs), so the cost of the subsidy to the government wouldn't change or may even go down slightly.
Yes, each such services are never going to materially change the cost to the government, however, if you could find 100 such services and they each make it that the cost of that infrastructure is covered by an extra 1 million passenger trips a year on each route which area an improvement then there could be enough of a change that it reduces the cost to the government (especially if it makes some longer distance travel more attractive as well - but reducing the risk of waiting a long time at an interchange location).
These examples might not be viable, however they're to illustrate the sort of options which could be looked at.
