Might need another destination to the south to get the trains out of the way (crayons suggest Abingdon).
Abingdon would be great but then you could also bring up Witney and design an entire Oxford commuter rail network. Cowley's actually funded which is why it's worth talking about.
Cowley is getting the Marylebone trains.
That's only 2tph which isn't much for a branch that serves some really busy suburbs and workplaces. The Cowley branch reopening would not be viable if it depended on traffic to/from London and Bicester - if you want the line be well used, it needs to compete on Cowley/Littlemore to Oxford station flows as well. There's a frequent bus connection at the moment but rail will be faster (especially at rush hour), more comfortable and higher capacity, plus offers great onward long-distance connections, it just needs to offer decent frequencies as well.
With a train every half hour, I suspect most people using the Cowley branch will be going to/from London, which will only be a few hundred thousand annual passengers max, far from the line's full potential. With the population catchment and both inbound & outbound commuting options to/from Littlemore and Cowley, half a million is very achievable, if not a million. For that it needs to be well served though. For example, in addition to the London services, a half-hourly connection to Begbroke, Heyford Park, Banbury etc will provide a 1-seat journey from many new housing developments straight to numerous workplaces (e.g. the Science Park) which would otherwise be done by car.
This should be the Oxford-Didcot shuttle
Not merging this with the London-Didcot service post-electrification would be a bit of a waste. Appleford is a small village but Radley and Culham would get a lot more than 100k annual users if they had frequent London services, as they're the main railhead for Abingdon (and still would be for southbound traffic even if Abingdon gets a rail link, as such a rail link would point towards Oxford not London). If the Oxford-Banbury stopper remains attached to it, it'd mean you'd have to use a BEMU or bi-mode which would be more useful elsewhere.
Crayons at the ready - wait for it - how about reopening the spur off the Cotswolds line and on to East West rail. Chiltern could revisit running to Kidderminster, but instead via Worcester. It might also take some of the pressure off overcrowding, where a heavily-used Cotswolds line train arrives in Oxford to a platform full of extra passengers also wanting London.
London-Kidderminster via Worcester would be incredibly slow. By being an extension of a London-Oxford express, the Cotswold line manages to be relatively competitive in journey times versus driving despite the slow speeds, so diverting it to Marylebone would seriously hurt London traffic. Ridership between London and Worcester, Malvern, Hereford and the Cotswolds put together is fairly significant as these are busy places.
Not to mention the fact that each of these train services (presumably 3- or 4-car) would use a path of a train that could be going to Oxford (could be 5- or 6-cars). The numbers just don't suggest it being a good idea, especially when the Cotswold line already has a rail service that's well used.
There's no convenient place to terminate a London-Oxford express every half-hour (as Ely/King's Lynn is for Cambridge) and even post-electrification, an extended Didcot stopper wouldn't take over much of the London-Oxford traffic, so the question of what to do with London-Oxford fast trains is valid point to raise, but screwing over a different market is not a good way to solve that issue. Moving the current 5-car IET London-Oxford hourly express over to an 8-car Electrostar would on its own add an extra 100 seats an hour without any capacity works necessary.
