"It's wonderful how our Dear gb cavils over how many angels can dance on the head of a pin, but doesn't understand HOW you would use a nuclear thermal reactor in space for propulsion purposes."

VASIMR can use nuclear thermal reactors, solar panels, or beamed power. But I don't know how you can use nuclear thermal reactor for VASIMR without using a steam turbine to generate electrical power.
Thermocouples, but they aren't very efficient- nor are they used except by RTG.

Or are talking about NERVA type propulsion?
http://www.daviddarling.info/encyclopedia/N/NERVA.html
"highest power: 4500 megawatts thermal power
5,500°F exhaust temperature
250,000 pounds thrust
850 sec. of specific impulse "

It's got a lower ISP.
And the discussion is how to get to Mars very, very fast. Almost interstellar speed fast.

I'm not anyway against NERVA type propulsion, but I think chemical rockets work almost well or better- if you have gas stations. NERVA have nearly twice the ISP, but chemical rocket engine is relatively low mass compared to nuclear reactor.

Of course, water does split into Hydrogen, and gas stations could sell just hydrogen for propellent-
whether for NERVA or VASIMR. Or fusion too.
It just makes oxygen cheaper- which is good news for
people in space who need to breathe.

But chemical or nuclear thermal [NERVA] can't do well if you want to get to Mars in one month's time. Maybe 2 months, but 2-4 months is about most you can expect from chemical or thermal- unless fuel and the spacecraft [in space] is really, really cheap.

Of course, it's not critical to get to Mars in 1 or 2 months- I believe 4 months for crew is reasonable goal.
I do see promise in beamed energy, and it's another market for space power [whether solar panels or nuclear generated electrical power].

VASIMR:
"The Variable Specific Impulse Magnetoplasma Rocket (VASIMR) is an electro-magnetic thruster for spacecraft propulsion. It uses radio waves to ionize a propellant and magnetic fields to accelerate the resulting plasma to generate thrust.
...
Current VASIMR designs should be capable of producing specific impulses ranging from 3,000 to 30,000 seconds "
http://en.wikipedia.org/wiki/Variable_specific_impulse_magnetoplasma_rocket

"Electrical power sources for the VASIMR engine could include such things as a nuclear power system or solar panels. For long-range flights, Chang-Diaz said, the best option is nuclear power. "Nuclear power is definitely a must if we're going to go to Mars," he said.
http://www.nasa.gov/vision/space/travelinginspace/future_propulsion.html

Though perhaps your point is get rid of the waste heat with a secondary type engine instead of trying to use radiators.
Probably, still need fair amount radiators though.

"Next he'll be proposing we set up "water stations" in orbit to feed all these steam turbines..."

I suppose nuclear reactor powered turbines could use something other than water. Maybe they could use CO2. CO2 isn't very common on Earth, but Venus is full of it as is Mars.
But yeah, human are creatures of water from a water planet, and water has many uses, so getting water available in space seems like a good idea- other than simply being something to make rocket fuel out of.