Define "return" though. Immediate financial return? broader tax returns? social value?
As we do with BCRs at the moment - we compare the "one off" cost of something against the benefits that it would generate over a twenty five/ fifty year period (as well as any subsidies required to maintain a certain expected level of service - e.g. a loss making line like the Tweedbank route)
That means comparing the impact of generating jobs and businesses/ boosting land and property values/ removing cars from the roads/ removing buses from the roads (and also the costs to existing train passengers - whether that's slowing down an InterCity service to stop at Reston or the lost revenue as some existing passengers migrate to a new station and therefore buy cheaper tickets for shorter journeys, e.g. we might sell slightly fewer Berwick to Edinburgh tickets as some of those passengers buy Reston to Edinburgh tickets instead)
Special attention can be paid to certain things that are "solved" (e.g. a train from Tweedbank into Edinburgh might remove a certain number of cars from the congested Sherrifhall roundabout on the Edinburgh bypass, so that might count higher than just removing a "typical" car KM
"Social value"? Sure, but I'm a bit wary of unquantifiable concepts since a lot of the time on here people use things like "social value" to cover over schemes with horrendous business cases, the kind of "yes, the average train would have fewer than a dozen passengers on board but there's this other benefit that magically makes the scheme worthwhile"
If we were truly rational then we would develop the biggest cities and bulldoze places like [insert post-industrial wasteland] into the sea... but we don't, and once you have accepted that it's not politically possible to make those decisions, then the process is already not truly rational and warped according to (small p and big p) political realities.
The same thing applies to regions, some regions because of their geography/demographics/industrial base etc. are going to produce lower returns. How do you sensibly decide what that region gets?
I do agree that making the final decisions at a local or regional level would be a good thing, but then you get into the need for regional democratic representation, which is a whole other wormhole to go down
I don't think it'd be "rational" to bulldoze existing places into the sea (albeit I say this living in a post-industrial wasteland myself, so maybe I would say that!)
However, there's an argument to be made against building (more) unsustainable housing, the kind that will cost a disproportionate amount of money to serve (public services etc) - in the way that I don't think we should be building more retail parks far from public transport etc
Some regions are bound to produce lower returns, which is why mass transportation is more suited to focussing on linking larger/ nearby urban areas than to provide the kind of rural links between smaller places some distance apart that members of the Crayonista community seem so fixated with (Tavistock to Okehampton, Matlock to Buxton etc) - one Crossrail journey would carry more passengers than a day of services between Aberystwyth and Carmarthen or between Dumfries and Stranraer etc
We don't have regional parliaments for everywhere but we do have Regional Development Funds etc which might be a good place for such things to be decided (until the UK embraces more devolution)