Based on the notion that there may be paths now for this service but none next year, does this mean that the Lumo London-Carmarthen service is planned to take over the path used by this service on the SWML? If so, the initial representation of open access services as capacity-boosters using up unused paths (which does have some merit, particularly on a stagnating GBR network) is starting to be somewhat less accurate.
I'm sure a London - Cardiff fast service will do very well (at abstracting increasing revenue), with the extension to Carmarthen possibly increasing usage at stations like Llanelli and Gowerton. However I'm also fairly confident that a TFW service on the same corridor, which does actually have a sensible stopping pattern and increases capacity on a high-potential regional growth corridor like South Wales - Bristol, will do much better, even if the requirement to cross-subsidise loss-making lines does result in no bargain fares.
Of course if both can be fit in then you may as well do so, as it's unlikely there are plans to use the space capacity for much else, but between five Ryanair-style London expresses or a regular frequency regional connector on a high-potential corridor, I know which one I'd prefer. It's true that any train to London does usually have the potential to generate much more revenue, but that's mostly because the network is so London-centric, growing other regional routes will help reverse that to some extent.