Quick update.... After my side-trip into Blender (well worth the time... great tool but the learning curve is gonna be steeeeeeep), it was time to start another TSRE sprint.
Turns out my four-headed monster wasn't as prototypical as it should have been for 1970, so I took a couple steps backward and had one of the UP managers I know who started with CNW give me a crash course on CNW signal aspects. That turned into a fun experience learning about how to implement negative states in signal scripts, and how you can't put a signal link into the negative side of a junction...
What drove that is interlockings... On the CNW, regardless of the number of diverging routes available in an interlocking, there's only one signal aspect shown for approach diverging or diverging clear.
Code:
A-------/2---------5\-------------10/----------------11
B-----1/----3\-------\6-----/8----9/-----12\---------13
C-------------\4----------7/--------------14\--------15
Using one head per path (1-2, 3-4, 9-10, 12-14) was the way I learned signaling ten years ago, and I had to unlearn it to follow prototype... to work with ORTS, I had to link the second head to the last possible node on a straight, non-diverging path, and apply a "not RouteSet()" condition in my script.
In this case, for Track B, the second head would be linked to 12-13. Simple, right? Not quite. MSTS signaling logic doesn't allow the back-side of a switch to be linked, so for track's A and C the links have to go inside two junctions i.e. A = 5 to 10 and C = 7 to 14.
Anyway.... Signal scripts seem to behave, signals have been placed and linked. For now. I'm sure I'll discover something just as soon as the next release goes out.
With a workaround in place, things are more or less signaled prototypically on both mainlines from Rochelle and Harvard to downtown.
I've also done a bit more track re-alignment along the New Line between Deval and Bryn Mawr, including the Des Plaines coach yard (used for MOW these days). Screen shots to come next week.