I want to be clear--I've been enjoying the show. Many people who know their stuff have very reasonably objected to this that and other aspect of it, notably the business of sending Space Shuttles all the way to Lunar orbit to rendezvous with the LSAM. (The LSAM itself is also pretty controversial, for good reasons). But until Pathfinder, while I agreed with critics that we saw on screen was not the most probable way the ATL NASA would achieve its goals, I didn't think it would take terribly large retcons to make it possible. (If only the shots of Orbiters we had in LLO happened to show they had an external fuel tank attached, one smaller considerably than the ones they launch with would be fine). Pathfinder is the first thing which, as shown, is practically impossible, and I wanted to stress too--if by some miracle of Yankee ingenuity the damn rocket were actually possible, it would have a hell of a lot more applications, and probably the scheme of using one spaceship one launches from Earth to go all the way to the Moon and Mars, instead of developing a deep space spaceship that doesn't do atmospheric braking and launches from a LEO space station and returns to it, is a silly way of applying it to the deep space mission too. Though given the rocket, it does have its points, just as I think I can defend the idea of Orbiters to Luna (on the grounds that knocking the final delta-V phase of braking off 3 km/sec with rocket thrust to return to LEO is overall more costly than using Earth's atmosphere to enable a heavier vehicle to be sent from Earth to the Moon and back again, and either of these reuse strategies are better than expendable one-way cargo delivery--or anyway, given we do want some mass to return to Earth, the reusable route is worth developing).
Anyway it is more plausible than say Space: 1999 or U.F.O,, and for that matter I can only enjoy Star Trek by ignoring much worse "physics breaking." And while things in the Gerry Anderson stories were most definitely physics breaking, generally speaking one does suppose "engineers can do almost any damn thing, given vision, money and time!" It isn't so much physics breaking as engineering leapfrogging, and again, note that there are other applications for this amazing invention.
No, we don't actually need fusion to explain it; fission would be no more problematic and less so in some ways.
I don't know how you define a "torch" drive, but I revert to those old Robert Heinlein stories he wrote in the 1940s. It is clear enough Heinlein had a bit of a loose grasp of the concept of nuclear power, seeming to think that total conversion of matter to energy would be just a matter of time, and not worrying too much about what the resulting photon rocket would look like in operation! Another benchmark for me is Larry Niven Known Space stories where the Belters used continuous fusion rockets.
If we could first of all cause nearly 100 percent of a given flow of fusible nuclei to actually fuse, and then harness all the energy that released to drive their own daughter products with that energy in a fairly well collimated rocket stream, we'd have what I would agree would be a Heinlein-Niven "torch" drive good enough for Solar System work, though Heinlein as noted elided it over to total conversion that would logically lead to a photon rocket (or in a version that is kinder and gentler, a neutrino rocket maybe). Whether we can use it for a slower than light starship that approximates light speed and thus gets the benefits of relativistic time contraction, depends on whether it is fusion or total conversion (as we'd get with say matter/antimatter reacted 1 to 1). Sticking with fusion, the idealized exhaust would be flowing out at about 1/10 the speed of light, since fusion involves releasing about 1 percent of the rest mass as kinetic energy of some kind.
Say we could make deuterium and tritium fuse in a continuous, and complete, flow (or make it fuse with pulses, but the pulses come fast enough, say 20 times a second or more, that they look continuous to the human eye). Actually 80 percent of the energy released would be in the form of neutron flux, only 1/5 of 1/100 the total mass-energy of the reactants would take the form of directly energizing the daughter helium nucleus. These nuclei would thus be moving at 1/22 or so the speed of light, half the speed we'd get from ideal fusion. But say we jacketed the core where these reactions are happening with say enough lithium that the neutron flow carrying the lion's share of the power output served to boil the lithium, and we capture all that heat with a helium gas turbine a la the Serpent fission engine, and use the resulting electricity we generate to further accelerate the helium nuclei we produce with fusion--then we can get them up to 1/10 light speed.
Well we could if we could do something about the waste heat the gas turbine would have to output--in a Serpent engine, hydrogen is the heat sink; since we'd have to throw away any hydrogen (or water, or what have you) heat sink material we vaporize to sustain the electric generator, we might as well throw it into the exhaust stream and use the power to accelerate it too. The result will be much greater mass flow but much lower Isp; thrust is higher but we need to consume more total propellant to go anywhere.
So if we had an ideal fusion process, Isp would be 3 million sec! that puts even the gee whiz 6000 I estimate for Pathfinder deep in the shade. We'd be able to burn for a million seconds or so before the masses we use up as propellant approach the mass of the payload (this is what the "seconds" in Isp mean actually, the exponential decay time a given mass of propellant can suspend itself in 1 G). A million seconds is a good long time to sustain acceleration at 1 G in terms of Solar system travel times, which is what the Niven and Heinlein stories were talking about. If you try to use such a drive to push a starship up toward light speed though you have to sustain it for something like 30 million seconds, a whole year--this can work but you wind up using godawful amounts of fuel to do it, making a Saturn V Lunar Apollo mission look frugal. In fact given fixed minimum material masses for tanks and so forth, there are practical limits even with an ideal fusion "torch." Go over to total conversion by some miracle of engineering, turn all your propellant into neutrinos speeding away mostly harmlessly in a rough beam in one direction, and perhaps we can bear the mass ratios and achieve practical sublight interstellar travel, because now the Isp is 30 million seconds.
Vice versa, we might find we need 99 times the mass of the product of deuterium-tritium fusion in the form of hydrogen (plain old proton nucleus hydrogen-1, but in H2 molecules presumably) to cool the power turbine, and then the helium we produce to generate the power is just a fraction of the exhaust mass, which is mostly hydrogen--with Isp of 300,000 sec. Now we can't sustain boost at 1 G for much longer than one day before we are seriously changing the mass of the ship--but that still gets us fantastic delta-V to be sure, and we don't have a proper Heinlein-Niven fusion "torch" but we still have a rocket with 50 times the Isp of the Pathfinder rocket, and can do a lot of Buck Rogers stuff even FAM's Pathfinder cannot.
So Pathfinder, stupendous though its drive is, is hardly a "torch" rocket by any reasonable definition. It falls between somewhere and leaves lots of room for progress, and can easily be a fission powered thing rather than fusion. In fact it isn't the ideal efficiency fission drive--fission processes release about 1/1000 instead of 1 percent of the rest mass, and so if we had an ideal fission rocket its Isp would be surprisingly close to the ideal fusion--a bit under 1/3 the Isp or about one million sec! Again Pathfinder's drive is far less ambitious than that. And I hope, a hell of a lot cleaner-I don't know about you, but I much dislike the idea of an ideal plutonium fueled fission rocket blasting fission daughter nuclei through the System. (To be sure, they would be moving way faster than escape velocity and soon make their way into interstellar space, and of course the mass flow would be low. Call me crazy then, I just think it is messy as hell).
Would it be more workable if Pathfinder refueled in orbit off screen from a big water tank launched by a previous Sea Dragon mission, before thrusting for the moon?