You got some ripening to go. Those hairs are too long. 'Shorter' is an optical illusion as the calyxes plump around them. they do coil/retract after dying. In my experience, most should be in this state when you see even 10% amber overall while scoping multiple vertical levels of buds... the top colas will have more.
trichomes coloration, pistils (% color and % coiled/dead), density of buds... use all three. Be consistent. You can then compare different harvest points better and make a more informed decision as to your preference on the matter. Do a few harvests at same point and form a good baseline, first. Otherwise it's just blindly stabbing in the dark.
I prefer the plastic vs glass reference for trichomes.
there's a lot of info out there in the guise of 'science.' You most definitely can relate potency and ripeness to trichome maturation and appearance, but what is best point? that isn't certain no matter what someone claims. Amber is definitely caused when cannibinoids breakdown into CBNA. THCA is mostly colorless and odorless, so associating any colors to it would only be correlative in nature. So, amber is most likely beyond the most potent it can be, but you'll want at least some amber to maximize overall yield -- your preference here may affect what you deem best. Maybe you's rather maximize the top colas and harvest lower colas a week later, for example?
I saw this related, recently, that relates very well to this question:
https://www.sciencedirect.com/science/article/pii/S2772375522000764
fig 6-8 are all very interesting... most of them from that point on in results section. Definitely hit up the discussion (conclusion? whichever) section too. Can see how genetics will preclude the possibility of a one-sized-fits-all answer. Maybe ADT or density of trichomes (seems the same thing to me, lol.. Average Distance nearest Trichome seems a oversophisticated way to reference density) will be more correlative ethan anything else?
Too bad they didn't measuer potency.. but you'd need varying harvest points for that which would have precluded necessary data for the hypothesis of this particular experiment. 1 slice at a time... then repeat enough to cover genetic variety, lol... it slowly paints a picture of reality.