There's still quite a bit of variety in the TNG era, mostly with short trips. It is with these insanely long intra- or intergalactic distances that we typically get both the travel time and the distance in hard numbers, but this is countered by many instances where we get a warp factor and an implicit travel time. Say, VOY "Scorpion": Borg intrigue 5.2 ly away, warp two for a hop that doesn't take hours of plot time.
It is pretty easy to argue that it's over long distances and times that the warp speeds "even out", both because local conditions grow statistically less relevant, and because there is good statistical evidence of what is a safe top speed for long duration travel yet always an option to ignore the limiters and lock the safety valves during short hops.
Considering they already considered for TOS that the main and the engineering hull ("star drive section") could occasionally operate separately
Did they? All dialogue references are to discarding
one component or another, rather than to continuing to operate it after a separation. It seems that a discarding sequence leaves just the saucer section surviving to tell the tale - and the propulsive capabilities of that section are never made clear, save for its supposed ability to break from orbit and escape the episode's calamity that is dragging the rest of the ship to a fiery crash.
As for the placement of putative reverse impulse engines, I think it's a bit of wasted effort. Only the various Enterprises
look like they could be moving on rocket thrust from the aft-facing impulse engines anyway; the other ship designs have their aft-facing engines located far away from the supposed thrust axis, meaning they would just start spinning in place if thrust were applied - or would have to fly in a pronounced nose up or nose down orientation we do not observe. Something more exotic than directed rocket thrust must be involved there, which in turn makes it unnecessary to have rocket nozzles pointing forward for reverse thrust.