I agree with your opinions on the Reading to Oxford service reductions, but I'm still far from convinced the new 17:23 to Moreton-In-Marsh will be anything other than a busy train with perhaps a handful of standees as far as the first stop on an average night. In other words, no different to many services at that time of day. It was a 2-car Turbo up until far more recently than 20 years ago - until late 2012 when the Class 180s began their second stint if memory serves. Yes, it was a busy train, but seats were usually available for everyone after eight minutes when it made its first call at Hanborough, and it left a little later then (around 17:32). With the new 17:45 hot on its heels from December I think enough people will spread over those two trains, not having to bust a gut to get down to the station for 17:23 or be left with the long wait until the next service that there currently is. We will of course find out in four weeks time.
Should I be wrong and it becomes too busy, there could be an easy fix for the following May timetable change (provided a 3-car Turbo can be found, which should be easier if, and it remains an if, 769s arrive on time to their revised delivery schedule), without messing around with local door operation. You make it into a 3-car Turbo and just take out the Combe and Finstock stops and put them on the 17:45 instead.
Ascott and Shipton calls remain on the 17:23 with 3-car Turbos OK to call and open all doors. The 17:45 has a 2-minute pathing allowance between Charlbury and Ascott (waiting for the 17:23 to reach Moreton) which means stopping it at Combe and Finstock could use that allowance and it would not need to have its schedule messed around by more than a minute or two at most west of Ascott. Also, with two less calls, the 17:23 would clear Ascott a few minutes earlier so that would help to give a fraction more resilience to what is a very tight timetable.
I know it was a Turbo a lot more recently - I had to try to get a seat on the thing out of Oxford many times during the period the 180s had disappeared and remember people lining up at the points where the Turbo's doors would open well before the train had arrived - usually before the connecting HST from London had even turned up - and precious few of those travelling from Oxford were/are going to the western end of the route.
It often left a lot later than 17.32, due to the chaos on the platform with the often ridiculously slow dispatch of the empty HST to the sidings and the passengers from Oxford mixed up with those connecting from Reading and London, plus everyone waiting for the northbound 17.36 XC departure - resulting at one point in the infamous 'stop all the Turbos way up the platform, out in the open, whatever the weather' edict, so that the Voyager could use permissive working to get into the platform behind it to try to minimise any delay to that train.
You seem to set a lot of store by the new 17.45 off Oxford, but, as I noted above, this is a peak train for ticketing purposes out of London, so the 16.20 to Oxford, with a change there, is going to remain the last option for anyone in London with an off-peak ticket to get to the stations out to Moreton-in-Marsh until the 18.58. Plus there will be the connecting passengers from somewhere beginning with R, who currently have a direct journey.
Please don't tell me you think that people will not be well aware of the connection, even if the pdf timetable doesn't show it - and I have my doubts that many passengers from Oxford who can make a 17.25 departure at present will be switching to the 17.45, whereas I do expect that train's introduction to make a notable difference to the numbers catching the 18.24 departure from Oxford, when compared with the current 18.17. There are a lot of people who are not able to get a train at 17.20 to 17.30, but could catch something at 17.45 or so and will do so like a shot.
Hence why I doubt a Turbo's ability to cope at 17.23 - if a five-car IET is sent out instead of a nine-car on the current 16.22 Paddington to Great Malvern, people are still standing out to Charlbury and every inch of floor space in aisles and vestibules is filled with standees out to Hanborough - who may not be overly enamoured of being expected to stand wedged in a vestibule every afternoon, whatever their journey time may be. Saying 'there's another train with more seats in 22 minutes' is not necessarily going to go down well.
The distinct impression I have got is that three-car Turbos freed up by the eventual arrival of 769s at Reading depot are going to be sent straight to Bristol, with nothing much more than two-car sets left at Reading, nor have I ever detected any enthusiasm for doing things like splitting calls at the halts. Having to serve them has been seen as a nuisance for a very long time now - and I can't see GWR wanting to use a scarce three-car Turbo, should some survive at Reading, just to serve two of the halts, then park it in a siding at Moreton-in-Marsh for an hour, with it finally getting back to Oxford at 19.50 after running back near-empty all the way - when it could have doing something else to earn some money in the meantime. Fit any three-car Reading 165s with the door controls the Bristol sets have and you might be talking.
As for putting calls at Combe International and Finstock into a 'Superfast' train, I don't think so. Mark Hopwood's phone would be ringing off the hook with calls from Worcestershire's MPs.