Regarding the former, it's been my thought that, as Shaw once suggested, each of the pressure compartments would be built separately, then assembled in drydock onto the main endoskeleton and the hull built around it, essentially creating a double hull both where compartments meet and around the ship's exterior where the outer hull covers the inner hulls of the compartments. (This would be what could make crazy refits like the Enterprise refit of TMP even plausible.) Unless I read incorrectly, I believe you think roughly the same.
You've read it correctly. Essentially the problem supposedly in older 2270s refit ships by the 2280s is that the structural stresses on the skeleton were causing problems with compression and structural frames. Hence putting trussed bracings in the pressure compartments is meant to help relieve some of that pressure.
Your interpretation of the trussed macrofilament model is also more or less correct, though the TNGTM also suggests the frames are then welded with the outer hull members into a combined stress hull with the interior being suspended by a separate lattice work that has substantial shock flexibility.
Using an automobile comparison, one might consider the old system the ancient American loved body on frame construction and the newer system being a semi-monocoque design.
Regarding newer ships being built using the new technology, is it your thought that these ships would completely abandon the hull pressure compartment model and essentially have completely adaptable interiors, or would still retain a pressure compartment design but be augmented by truss SIF fields?
I'm a little torn on this one. I think this would be a gradual evolution, but that a Miranda
constructed in 2310 vs 2330, the latter would no longer be using the hull compartment model. The interiors we see of TNG/DS9 era TMP design ships suggest that these have interior commonality with modern Federation ships. (See: Saratoga
from "The Emissary" or albeit non-canon, the Righteous
from "Star Trek: Borg")