I'm also not sure that THC-A is exactly attributed to resin content? Donkey Kong doesn't usually appear to be my most resinous or frosty plant. I think my Grape Sromper is the frostiest. She only tests around 24%. Looks like it should be 34%. DK does have some huge sand grain type of trichomes and the scissor hash will dab pretty clean on a hot nail. I can tell you in many ways that this pheno of DK, as a strain is an anomaly. Not only in it's tested potency. It's easy to grow, stacks and structures like a champ, huge yields and quick finish. Can't forget the KO high.
What are the trich heads like on your GS?
If they're smaller than the sand lot trichs from DK that could explain a lot. The physical structure of trichomes is a ball on a stem. The bigger the ball the more volume it contains. So it would stand to reason that smaller trichomes not only have less volume, they have more lipids as well containing that smaller volume of oils. This alone would skew test results to favor large trich heads.
Biologically speaking the light does not directly increase resin production, cannabinoid synthesis supposedly occurs during the dark cycle, so the light would directly influence flower production moreso than resin production, which alone will drop your potency a relative 2-3%? So your 34% thca would be (34 x 0.97) not accounting for the degradation of resin glands as a result of more intense light. If I'm correct, in theory the same strain under de's should see an inconsequential drop in cannabinoid but a sizable increase in yield
IME, MH lighting seems to increase trichome content while limiting the amount of plant material being grown, when compared to HPS. Of course, I'm speaking from rudimentary observations over the last few years growing.
I've no experience with DE or CMH fixtures but am definitely curious about their performance.