Their only benefit would be portion working, and there is hardly any of that going on. Other than that, and other than a few services like the Cotswold line where some 7 car units may have been of benefit, a double 5 car costs more to buy and maintain than a single 9 and costs more to staff properly (a trolley and first class steward only in half doesn't count) because they aren't gangwayed.
Those have a better case because they actually do portion work.
I don't to be fair support the "occasional extensions" model. I prefer the approach involving a clockface timetable with planned connections, so a quality connection is available every single hour (or more), not an odd through train at a time of day that may not be convenient anyway. Keep it long, keep it frequent, keep it simple - the basis of Takt.
Taktfahrplan may work in Switzerland, where they have planned and invested for many decades to make it possible, plus they do not have British politicians and their attitudes towards public transport to contend with - and German attempts to achieve something similar remain mired in problems. As for France, off the TGV routes many of the timetables do a passable impersonation of BR services from the darkest days of the 1970s.
And what pray tell, do you suggest for the Cotswold Line? Because the London-Worcester/Malvern/Hereford operation it is not an occasional extension of a service, is it? But we are not going to get a hourly nine-car train each way, plus extras in the direction of the main peak flows, as it would be utter madness financially when it comes to operating costs.
Why is it that people who live nowhere near the Cotswold Line are so keen to keep volunteering it for seven-car trains? If GWR thought they would be a good idea, it could have made a case to the DfT as part of the 802 order, but it didn't. They would be too small for the busiest weekday peak and weekend Cotswold Line (or Cheltenham) services, too big for many of the off-peak runs and too small if something was needed to step up at short notice for a busy Bristol, South Wales or West Country duty.
The ability to step up the frequency of the Carmarthen service in May* rests entirely on availability of five-car IETs and portion working west of Swansea. Without this option being available (and the potential open access competition, of course), GWR probably would not bother, as no one is going to pay for the infrastructure work needed to fit nine-car sets in Carmarthen station. *Assuming, of course, that you accept six trains to/from London plus a pair to/from Bristol Parkway count as more than 'an odd through train'.
If commuting had not been hit hard by the pandemic, I have a sneaking suspicion that not quite so many nine-car sets would have been available at present to operate London to Penzance duties. Further changes to travel habits may well mean GWR has to do more juggling with which sets work where and when in years to come.
The LNER Lincoln-London service would probably not wash its face financially at the current frequency if worked by nine-car sets - and given the sentivities around any extra train movements in the Newark area, I cannot imagine that accommodating a clockface connection to/from Lincoln is top of the wishlist of Network Rail or the train operators there.
