In the case of Super-glue, Gorilla glue, or two-part epoxy I would agree with the "Chemical" bond theory; but in the case of all PVAs from Elmers to Corner Weld there just is no "Chemical" activity.
PVA is inert. That's why we can use it as a conservative glue directly on some items.
At the risk starting a firestorm, David, you need to go check your math. 35% and 35% do not add up to 100%. But aside from your math on a mitre joint [50%end-grain and 50%long-grain] what matters here is the surface area of the long-grain joint.
If the joint is on a 2" wide moulding 1" thick then the two surface area are disparent as such:
Mitre: [given: the 45 mitre surface on a 2x1 is 3x1] so for a 50% long-grain surface on said mitre is 1.5 square inches.
Cope and stick: [given: tenon thickness is 1/3 the thickness of the moulding/board] so for a 2x1 the surface area will be [using your 1/2" long tenon and mortice] 1/3"+1/2"+1/3"+1/2"+1/3x2" for a total bond surface of 4" which is 100% long-grain bonding, and too small to be affected by seasonal migration, which allows your children to "swing" on the doors.
The reason for a mechanical bond [v-nails or spikes and brads] at the time of gluing is to draw the bond together into a greater than 1lb tension [glue in mitre vise than nail when dry] and less than 50lb tension resulting in a starved joint [no glue].
A PVA bond happens mostly from surface tension and very little "tooth" or penitration into the grain. Which is why all joints (that can be) in furniture (that receive incredable stress) are planed to a glass smooth joint. [I'm speaking as a woodworker who makes furniture with hand-planes as appossed to IKEA.]
What penitrates is the water in the glue; only carrying a fraction of adhesive. This results in a thin layer between the wood surfaces that has filled any tiny nooks and cranies and resulted in a shear bond that is a result of surface tension.
The V-nails serve two purposes. 1) to hold the bond in place while it drys (it takes MANY hours for the moisture to "wick" away from the bonding wood) and 2) if the glue bond fails the nails or V-nails will hold the frame "whole" whilest the customer rushes (usually within a year or six) it to a restoritive agent (framer).
IF you are in the habit of gluing in a vice then V-nailing, be aware that any shock or jarring to the still moist wood and glue bond will result in the degridation of the bond that may or maynot be evident right away.
So spread a nice thin coat all over all surfaces to bond, then V-nail with a minimum amount of nails neccessary to hold the bond correctly and let rest at least over-night before fitting.
If you doubt me, as for the strength of a glue joint [long-grain to long-grain], simply take two blocks of wood 3/4" thick, 2"-3" wide and about 6" long, plane a glass smooth joint (or sand to 600 grit), apply a 2mm dot of PVA every inch on one side, push the blocks together and slide back and forth untill the bond takes place... now, without rocking or bending try pulling them directly apart.
Now put them back together, re-establish the bond by sliding them back and forth untill they resist. Set them aside for a few days. Now try to break the bond. Trapper, you can use your mallet.
As for nails rusting, v-nails falling out and glue joint opening up on the inside... these are all due to the drying out of moulding that started at more than 4% humidity levels; NOT seasonal wood swell and shrink (migration). And the drying out will not occur more pronounced on the top half because of hotter dryer tempature in a room. The moisture is IN the hotter higher and dryer near the floor, so the frame drys equally. [Thank you Getty Museum].
BTW: Norm Abrams is a carpenter and is still learning how to build furniture. But he is better at promoting woodworking than Martha was at framing; so we woodworkers still love him.