What we’re talking about I think - to be quite vague - is a technology that best converts between energy and matter. The more of the latter you have, the more effective the propulsion system will be in this sea of gas and gravitation. Sound right?
Generally, a rocket engine is more effective in terms of fuel used when it puts more energy into less matter. Energy means kinetic energy. Thus, an efficient rocket engine has higher exhaust speeds for less matter exhausted. For atmospheric engines things change a bit, since air may be taken in and used as the exhaust. But basically you still want to use the least amount of fuel.
Now for FF atmospheric propulsion, one would probably have to send the FF ion jets into a heating chamber, where they heat the air, which is then expelled to the back. It is much like a turbojet or a ramjet engine, only you use the FF exhaust to heat the air instead of burning some petroleum fuel. In terms of fuel used, this is much more efficient than a jet engine, since you use nuclear energy and the fuel quantity is some 1 000 000 times smaller. It remains to be seen whether it is more efficient in terms of engine mass, too. You don’t want a huge, heavy engine to do the same work as a jet engine. If you wanted that, the solution were right around the corner: a thermal nuclear fission reactor powering an aircraft engine. Think about an aircraft carrier power source on an airplane.
Nuclear propulsion has been tried:
The main issue that it wasn’t continued was political, due to fear of a Fission reactor crashing in a populated area.
The main technical issue was weight, reactor + shielding being quite heavy.
Like most technical issues there are solutions to these problems, but there are obviously practical limits.
For FF, the main issue for space use in an orbital launch capacity is you are effectively trading weight. Yes you have less actual fuel, but the reactor itself has a large mass and your propulsive element still needs to be there.
Air with a turbine is only practical up to about Mach 2.5 at which point the air temperature entering the engine is so high that it literally melts the engine.
Ram systems can work to higher speeds, but they are not self starting, so you need multiple propulsion elements. Using an electric heat source would require the system to generate ridiculous temperatures as you need to maintain sufficient deltaT to generate thrust. See SR-71.
Above about Mach 5-10 you’ll need internal mass to throw out the back as the temperatures will well exceed the limits of the materials we have today.
Starlite was a possibility, but unfortunately Maurice Ward died last year and as far as is public, he took the recipe with him.
As you state more speed/energy can result in reduced mass, but there is a limit, in the sense that, above a certain temperature you’re effectively throwing a hugely powerful ion beam out the back of the engine. This would (at least) create the same political issues as with flying a Fission reactor, if not more.