And only comes into play at warp speeds. Under impulse power it isn't all that useful; ordinary shields more than suffice.
I'd argue the exact opposite: at warp speeds, the discontinuity at the warp field edge will already help in protecting the ship from the "real world", while at high relativistic speeds the ship is in real danger of being pelleted by deadly dust and heavy nuclei.
Certainly the deflector is up and running at sublight flight modes, and is specifically lit up for that in ST:TMP.
The simple "good reason" is that the impulse engines don't need to be anywhere
That's not a good reason. There would have to be a definite reason for
not mounting the glowy bits on the thrust axis, the default position. There never is.
If thrust vectoring is the primary mode of operations, then the glowy bits could be at dorsal bow; the rear mount makes no particular sense in that case.
under impulse power there is again no special reason for a preferred orientation
The Trek fact of the matter is that the ship always moves bow first, with bow defined as "opposite the glowy bits of the impulse drive", but with the glowy bits oriented whichever way (often tilted upwards).
There's no point in exploring the nature of the impulse engines from the point of view of hypothetical starships that behave in a manner not observed on screen. The observed behavior leaves two options: it's not a rocket at all, and thus goes a long way in explaining why starships don't behave like rockets at all, or then it's a thrust-vectoring rocket (perhaps among other things), only engineered in a truly perverse manner that makes one think it's a system of tertiary importance at most.
If you can travel at warp 1.1, you can travel at warp .1 with the same engine.
There's no basis for such a claim, just like there's no basis for the claim that a rocket engine can throttle through all imaginable settings. Some special rocket engine might do that under certain circumstances, but in reality, none do; the reality of warp drive probably reflects that as well.
The division to warp and impulse engines is a step towards realism; integration of the two distinct drives would realistically result in compromises that diminish performance in both warp and sublight regimes, and may thus be done only as an emergency measure (that is, a frigate today
can maneuver at port with her turbines, even though the diesels are the reasonable choice) or as a cost-saving measure in a compact vessel.
Perhaps (on the TOS-E) the centreline positioning of the impulse vents is also where the emergency (newtonian only) thrusters are located
For all we know, the things we mistook for impulse engines
are the emergency thrusters, and the real impulse engines are the glowy spheres in front of the nacelles, or the subtle grey squares on the pylon surfaces. The TOS ship is ill defined in that regard. Indeed, no Trek ship has had dialogue that would specifically establish where the impulse engines are, or VFX/dialogue synergy that would indicate their exact location through some "hey, now the impulse engines activated and lo, these red things here started glowing at the same time - perhaps there's a connection?" manner.
Timo Saloniemi