So, even here in a TNG episode, we have ambiguous dialog that could indicate that that "starboard nacelle" and "starboard antimatter pod" are two terms for the same thing, at least in this case.
To me it's really not that strange, at least in TOS: If my engineering proposal is accurate, the "antimatter nacelles" really
are the antimatter pods, since the volatile stuff is not stored anywhere else. So for the crew to use the terms interchangeably is perfectly understandable, and even segues into the TNG terminology (where on starships at least, the pods and nacelles are separate) without too much of a stretch.
Mytran, that sounds good but I think it is missing the central or main reactor component.
As I hopefully demonstrated upthread, there's really no need for a central reactor
There are oddly instances where the Enterprise needed to refuel her impulse engines ("Doomsday Machine", "Mudd's Women") but she appeared to lack that ability and it wasn't due to damage.
Enterprise's seemingly limited range on Impulse in DM has always been an oddity. But I wonder if it's not somehow connected to the Warp drive. After all, the big disadvantage suffered by the Ent in that episode was that the antimatter had been "deactivated". What if that affected the Impulse engines too, because what if the Impulse reactors have their own antimatter supply that is used to "spice up" the energy output? Coupled with a (mass reducing) low level subspace field generated by the nacelles, and there you have a drive system able to propel a ship at speeds of 25% of LS with relatively little fuel consumption. Knock out the antimatter and it becomes a traditional newtonian rocket engine, which at full pelt will exhaust it's fuel in a matter of hours (especially if called upon to provide shields and phasers). The same kind of problem may explain the situation in TNT, except that time of course it was a drunken Riley causing the problems (I'm also not discounting the possibility that he may have switched off the Impulse reactors as well, which may have required an even longer restart procedure. It might also be why the refit-E seeks escape from the Reliant at such appallingly low speeds near the end of TWOK - no subspace field, since the mains are offline, so they are flying on solely newtonian momentum.
Channeling "all available power" to the Impulse Engines is still possible in this scenario; as I said, I envision the dilithium crystals being a power converter (instead of a M/AM reaction mediator) to energy from that can be easily diverted to where it is needed, be that shields, weapons or impulse drive systems.
Based on the above events, what if the impulse engines don't directly use matter fuel? What if impulse engines are powered by energy delivered from impulse power system that normally is charged up by the warp engines (dilithium power conversion) or like in TMP, directly powered by the m/am dilithium system. "Impulse power" then is just power stored in giant batteries and only comes into play as additional power or emergency power is requested. How this dependency plays out I haven't fully thought out yet.
I do like this original line of thought! It also matches what we see in TMP, but how well would this carry on into the TNG era though? The only reason I ask is that Geordi tells Scotty that Impulse Engine technology hasn't changed much in the last 200 years.