This is a seriously impressive post.
As someone who took up railway signal engineering as a career 53 years ago, I am humbled that someone from outside the industry has put this together so well.
Re your asterisked note, points generally do not auto-normalise. There are occasions where this is done for specific reasons, but it is rare, and there is no reason for it to apply here. (Your observation of niche is spot on!)
In addition, what this timeline shows is that comprehensive approach locking (which I commented on in an earlier post) was irrelevant here; the timeline would have been identical either way.
Well done to you, and well done to the driver and signaller, who appear to have managed this masterfully.