Have you had the ruby red extract analyzed? I've recently had my assumption that d9 THC goes to d8 and then to CBN, followed by other polymorphic degeneration byproducts loudly challenged on another forum.
The only analysis tool that I own that could even come close to identifying the degredation products is an old dual beam spectrophotometer that is UV capable. This has given clues what is happening a bit but this is not my focus to learn the precise nature of what is taking place. I do know the spectral absorption is indicative of known cannabinoids but not necessarily which one(s). The aborption peaks are pretty damn close between the different cannabinoids and there are potential techniques to sort this out but I am just one old guy lolz and it takes a bit to llight the fire to answer the next hard problem. Debate is fine but direct observation is more meaningful to me.
I have severe symtpoms that I use this stuff to treat. My current focus is learning how the differently aged samples of compound impact the symptoms. I am also intentionally using a "cooking" process to alter the purified THC which is to say a very high heat controlled boil to alter the compound. There is a HUGE difference in symptom relief between different stages as it degrades. One month at a time I try different combos and believe me when I tell you the symptoms/releif are a much more important study atm.
If you look at my channel you will see various degrees of color in the final compound. All of these are tested by me to see what impact they have. Then I move on to further refinement and sample. Then when it is nearly pure and clear as it degrades and I encourage it with heat I also take note of what helps me best.
I realize your question was seeking the logical molecular explanation but this is honestly a much tougher question to answer than it might seem. Since the sample is continuously degrading in a sense then at this moment it may test as one thing but wait a day and then what? It would take a long term wavelength absorption study of the sample on my UV equipped spectrophotometer in my opinion to even come close to knowing what is happening because the only answer that has meaning MUST involve the "integral" value - integral is time and it is time that integrates an answer like this. An LC seperation will help I suppose too.
That being said I very easily seperate the red from the clear from time to time and I do save the red compound(s) in a small petri dish. I am kind of lazy lolz but a run through my small liquid chromatography tube (just old school stuff) might give me an idea at least of the number of compounds present in the "red". LC testing is fun but it is a bit messy and I have gotten a serious case of too many dirty test tubes already laying around to solve before I proceed lolz.