Length of up through line at Swindon?

At a rough guess around the 250-300m mark? That's to the points which are by the station, otherwise a 550m freight shouldn't really affect the access, as usually a London HST would use the Rushey Platt relief, and then straight into 1 or 3. I believe the crossover speed is faster than the ones near the station.

While this information is more easily available to the respective developer or people with a tester's license, there's a little trick with which anybody can find out the length of a track section, at least as long it is visible on the panel.

Open the Incident Control Panel, head over to the Track Sections tab and use the "click to select" button to get the identity of the track circuit in question, e.g. TSUE for 65's overlap.
Now replace the initial "T" with an "U" (i.e. TSUE -> USUE) and head over to the TSRs tab. Click on the "Add..." button, switch to "Individual subroute" mode and look up the U... subroute you've determined previously.

When you select a subroute, you can then see that the From/to position and speed limit fields are populated with some default values that represent the respective subroute's actual length and speeds.
For track sections containing some pointwork, determining which subroute with its respective suffix (-AB, -BA, AC, etc.) matches which physical direction can be a bit tricky, although
a) line speeds might provide a clue, and
b) track section ends are supposed to be numbered clockwise from A to Z, starting at the 12 o'clock position

I your case, USUE-BC and USUE-CB are the relevant subroutes, which are shown with a length of 180 m, which is therefore the additional length you have available over and above the standage available between 42 and 65. (If the rear of your train would be overhanging TSUD as well, then 2403A points will remain locked normal and you're in the same situation as in your original post).

while points 2120 at Rushey Platt had 75mph. Looks like there would have been a difference after all.

Note that there's a small sim error in this area though, which means that the 85 mph on the Up Main (= 10 mph speed differential towards the Up Relief at 75 mph, hence the free green at signal 53) is currently missing in the sim and is shown as 125 mph, too.

I understand these are also track circuit lengths? Never knew you could do that!

When the Rushey Platt turnout to the Up Relief became a high speed junction, pretty much everything scheduled to stop was routed up that Relief - not the main. I don't recall when that was though. It's possible the timetable you used predated it, or was incomplete with line codes.

It never was timetabled separately, and signallers were free to use either option as they saw befitted the situation - although as eluded to upthread it was most certainly the faster route via the Up Relief.

As far as I know I've always used the Main while playing SimSig. I was honestly surprised when my GWR train took the Rushey Platt line when coming from Bath in October.

The train had to wait until its platform was free because of some disruption, so I thought that I had been looped.

b) track section ends are supposed to be numbered clockwise from A to Z, starting at the 12 o'clock position

I'd just taken the clockwise rule from the "Basic Guide to Sim Construction" overview on the Wiki, which must have been simplifying things a little. Thanks for the in-depth explanation.