The real issue of the OP seems to be scratched/pitted/ or dirty glass. He has found that large source illumination gets under the imperfections in the flawed glass, and eliminates them, but this is of course at the expense of any further control over any specific area of the print. #2 solution is to simply lay the negative on the paper, sans the offending glass, and hope for the best, because the math, absent any paper curl, negative curl, humidity, vibration, bump, or draft, says it works. (and it very well might, now and again)
I say get some glass that doesn't suck.
Contemplating the vectors of the photons probably isn't much use in the real world pursuit of a fine contact print, which is, as I understand it, the goal. Time spent doing is the way, and plenty of more accomplished and eminent foreheads have beat the walls before us. Contact printing is as simple as it gets, and brutally reveals any deficiency in the negative or the process. You can't calculate around that. Get some good glass, and clean the printing area- that's the problem, and you can figure all the day long, but it won't change until you do something about it.
I don't think any photographic pursuit, mathematical or otherwise is a waste of time, but I honestly don't see the application here, when what's needed are the proper tools, and attention to the details that matter.