Mike, Im a newbie relative to many of those posting here, but I have had experience of fitting points to a prototype plan (Helston GWR if you are interested - has just 8 turnouts five of which are curved on both roads and two of which more or less merge into each other).
What I did was start with a 1:40 plan I picked up from somewhere which, for good measure, I also overlaid onto the OS 6inch map (which you clearly have) so I could cross check them. I then scanned the the plan to a jpeg and imported it as a background into Templot. It was the first time I had ever used Templot and its not an easy tool, however the videos on the 2mm Youtube channel (which weren't available at the time) look good.
To start with its quite easy to get the crossing angle from the plan (or the 6in OS map) by simply measuring it, which gives you a good start. Once you have that its a lot of messing around with curve radii to get the alignments to work. This is a bit tedious but Id much rather be messing around with curve radii on a computer than building and rebuilding points which look right, but when you actually come to butt them up, don't quite align properly. To be honest I really couldn't imagine getting things to fit properly even on the relatively relaxed geometry that is Helston without Templot (or some similar tool) so its definitely IMHO the way to go if you want to come close to matching what appears to be a very complex geometry without having ugly kinks where transitions between components aren't quite properly aligned.
A trick I found useful with a series of points where both roads are curved was first to draw them as overlaid pieces of curved track. This helps work you to visualise the alignments between the various sections first and also shows up where the plan is probably wrong. Whether you then build a point over the top or change the curve into a point is up to you (I did a mixture).
It took many, many hours and several false starts to get a flow of curve that I was happy with and which was sufficiently close to overlaying the plan that I felt reasonably confident that it matched the prototype, but the effort was well worth it and, as I say, far better than building points which then don't align properly with each other when placed on the track plan.