How much more efficient is it? The fact that you have to carry propellant with you anyway seems like a big factor in favour of traditional rockets, where the fuel becomes the propellant. And burning the hydrogen way hotter seems like a big engineering problem - rocket motors already run at temperatures which pose materials science problems, increase that a few thousand degrees and you might as well say we should run the spaceship on fusion power.
I don't doubt that the physics works, and it's important to note that say a 2x gain in thrust-to-mass ratio would lead to enormous advances in trips out of Earth's gravitational well (10x? 100x?) due to the tyranny of the rocket equation. But I'm curious whether this is the 2x gain of replacing oxygen+hydrogen with hydrogen, plus a big complex reactor, or the millionfold gain of replacing hydrogen fuels with uranium.
Edit: TFA says twice the specific impulse, and presumably that's the optimistic estimate. But still very good!
NTR is not anywhere near 2x chemical because of parasitic losses, in fact it may only barely offer higher net DeltaV.
NTR requires heavy engines, heavy shielding and heavy radiators to keep cool. The final NERVA prototype was as close to a functional NTR as ever built, and it massed 40,000 lbs while only generated 55,000 lbs of thrust, with a maximum ISP of 710 seconds.
A SpaceX Raptor only has a Mac ISP of 380 seconds, but masses only 3,000 lbs, and produces 500,000 lbs of thrust. Add another 50,000 lbs dead mass to the NTR for shielding and cooling, and you see why Raptor will get humans to Mars well before any NTR and just as quickly.
TWR is only important when you have to climb up the gravity well. Once you're up in orbit, high ISP is king to get you anywhere. Other comments already highlighted that NTRs are a middle ground between superefficient, ultra-high ISP electric drives, and very high TWR chemical rockets. So that's where this sits: high enough TWR to be practical, but way better ISP than chemical (though still far from electrical).
NTR is not competitive with chemical rockets in any actual application. Dry mass has to triple at a minimum, your propellant evaporates over long trips, and NTR engines don’t have enough thrust to land on Mars or even the moon, requiring the additional dry mass and complications of specialized landers.
We need a huge step forward in NTR before it’s going to be useful at all.
Those already take the slow boat Hohmann transfer orbit to maximize payload capacity. Starship might be able to send 150 tons that way to the surface itself. And building bigger starships is easier than building NTRs.
This sounds like a matter of scale. A nuclear spaceship (a theoretical one, like all of them) with twice the thrust wouldn't be carrying twice the dead weight.
So at some level of scale this outperforms a traditional rocket, and your oddly impassioned argument about why SpaceX is so much better is just relevant to particular use cases rather than spaceship design in general.
Scaling up low thrust to weight engines with large cooling and shielding requirements doesn’t create many mass efficiencies. Maybe in the shielding, but that’s the least of your concerns.
The other problem is that low thrust to weight means an NTR spade ship can’t land on Mars, or on any body with a significant gravity well. So you need to bring chemical rocket landers, increasing your mass duplication and tech complications.
A multipurpose chemical rocket powered space ship Luke Star Ship is far more practical and nearly as fast.
I don't doubt that the physics works, and it's important to note that say a 2x gain in thrust-to-mass ratio would lead to enormous advances in trips out of Earth's gravitational well (10x? 100x?) due to the tyranny of the rocket equation. But I'm curious whether this is the 2x gain of replacing oxygen+hydrogen with hydrogen, plus a big complex reactor, or the millionfold gain of replacing hydrogen fuels with uranium.
Edit: TFA says twice the specific impulse, and presumably that's the optimistic estimate. But still very good!