That's compared to a few hundred pounds for an equivalent chemical rocket. The point of post you are replying to is absolutely correct, it only makes sense if the rocket using it is large and carries a large amount of propellant.
There are effectively two factors in rocket design - the engine ISP and the mass ratio. The mass ratio is a measure of how much propellant is carried VS the dead weight (engine, tanks, payload). Those two things can tell you the velocity change of the rocket (see: "rocket equation"). Note that you have to recognize that the ISP and exhaust velocity are one and the same to make sense of it.
The ISP is twice as good as a chemical rocket, but the dead weight is very high, too, so for this to make sense, you need a large amount of propellant. The difference between 250 lbs (chemical rocket weight) and a few thousand (practical lower end of the NTP and associated shielding, etc) could be critical.
For example, using hydrogen as a working fluid increases the ISP (the lighter the exhaust products, the better) but reduces the mass ratio because the density is so low, the tanks have to be gigantic and therefore heavy. reducing the mass ratio. If it used Xenon, it might have lower ISP but the dead weight would be smaller due to much smaller fuel tanks. It's a trade-off, and NTP engines don't care very much what fuel they use.
Someone has already figured all this out, there was a perfectly sound design for a rocket upper stage using a NERVA engine, I would suggest that as a point for further research.