Unfortunately this story is now down the page, so this probably won't be read much, but I'm going to correct the false assumption here that seems to have played a major part in this thread.

NASA does not launch military spacecraft. That job, today, falls to the United Launch Alliance (primarily, smaller payloads can go on other US commercial providers), a wholly separate organization from NASA. (ULA does occasionally launch NASA spacecraft, but at that point, NASA is simply a customer who is buying a ride to orbit.) The last time NASA itself launched a military payload was STS-53 in 1992. Since then, all payloads have gone up on unmanned Air-force or commercial launch vehicles. (Why is this? Challenger. The military did not want to be grounded for another two years if another shuttle had an accident.)

So no, we do not need NASA for national security, and have not since 1992.

Back to the point of the main article, I find it interesting that congress appears to be perfectly happy to send hundreds of millions of dollars to Russia for rides to orbit, but have to be dragged kicking a screaming to let NASA pay some American companies to develop the same capability, possibly for even cheaper (i.e. SpaceX's goal of 20-30 million per seat to the ISS)

Though a vehicle may be designed to work in 0.38 earth gravity, that doesn't mean it will collapse or otherwise not work in standard earth conditions. Most often the structural driver for spacecraft, rovers, etc is the launch vehicle environment. Curiosity will be going up on an Atlas V, which will subject the rover to 5-6 G and a strenuous acoustic, shock, and vibration environment. In addition to the launch loads, it also has to survive the sky-crane landing on the surface of Mars. So it really isn't too surprising that it can support its own weight on earth.

You are correct that fuel for attitude control is generally the limiting factor for spacecraft (useable) lifetime. Using solar sails for attitude control would be possible, I think, for spacecraft operating far enough away from a planetary atmosphere. Otherwise, drag from the sail would certainly overwhelm solar pressure. So, though it may be possible, I'm not sure how economical it would be to use for stationkeeping. I would be interested in seeing a trade study between electric propulsion (another low thrust over long duration type system) and solar sails. Solar sails would probably mass more and take up more volume than an equivalent EP system, but would not require nearly as much electrical power, which would reduce solar panel size and save some mass and volume there. Bottom line, though, is that with the state of solar sail tech right now, it'll be a while before anyone tries using solar sails in such a manner. Most solar sail applications I think you'll see is as the main propulsion system getting craft between planets. That's where the greatest benefit of low thrust, but long duration burns that don't require propellant or little to no electric power will be.

Aside from propellant, the big limit is degradation of solar panel generation capacity (from atomic oxygen, radiation, etc) and battery capacity that also degrades over time. Moving parts (like momentum or reaction wheels that control spacecraft pointing) can also wear out. Space is hard on lubricants. Then there are things that can ruin your day like computer errors from radiation effects (like the Galaxy 15 spacecraft) and impacts from space debris. So while eliminating the need for attitude control propellant will improve spacecraft life and is a worthwhile endeavor, don't count on getting unlimited spacecraft life any time soon.

One thing I haven't seen discussed is that once this rocket is built, what the hell is going to fly on it? I don't see any mention of funding for building a 75mT payload that will then be ready to fly when the rocket is. Right now, all I see happening is that this giant rocket will be built and then will sit around for years waiting for something large enough to fly on it to be built.
Which raises another question: Do we even need it? What's the point in having a giant rocket without giant payloads to fly? What I would like to see is a study comparing the cost and timescale of doing the following two options:
1) Build this huge rocket, utilizing the expensive legacy shuttle hardware. Then design and build a huge payload as justification for your huge rocket. (Sound familiar, sort of like the ISS-Space shuttle relationship?) Fly this huge rocket a couple times per year, making you unable to amortize fixed costs (launch pads, support personnel, etc) over more launches. If the launcher fails, you've lost 75 mT of payload. Ouch. So you'll also need to spend more money making extra-sure that it'll succeed. Oh, and if you fail, your mission is completely grounded until the vehicle is fixed and re-tested and certified for flight again, because there are no other 75mT capacity launchers in existence.
2) Start designing and building payloads that will fit on existing (or near future) commercial launchers. Start a market for even more launches. Let some economies of scale come into the picture and reduce launch costs for you. Let commercial companies compete and bid for your business. If the launcher fails, you've lost a lot less payload than in option 1. Inconvenient, but not as devastating as losing option one's super-launcher. Also, if one of your commercial launchers does fail, you have some more to choose from to launch things in the meantime while the failed rocket is investigated, fixed, re-tested and re-certified. Your entire program does not have to grind to a halt while the launcher is fixed.

Now, which of those options would result in more activity in space? Personally, I think option 2 would be the way to go. More opportunity for cost saving. No single point of failure to get your stuff into orbit. You can start designing payloads to go up right away, instead of waiting for the funding to become available after your shuttle-derived super launcher is ready.

But, of course, what I wrote above does not matter to congress. All they care about is: Does this program pay back my donors enough for them to keep supporting me?

What trade studies were done that decided a 75mT payload capacity was needed as opposed to a 50 or 60mT? Is there a linear increase in cost vs. payload capacity? Is 75mT some sort of optimum, balancing cost vs. development time vs. existing hardware capabilities?

Or is it a number pulled out of someone's ass?

Arguing requirement vs. design is mostly semantics at this point. What matters is where the number came from and what sort of analysis went into it (if any).

Likely the FCC because the vast majority of commercial spacecraft are communications spacecraft. That means the FCC already regulates power/frequency allocation for ground and inter-spacecraft communication for those spacecraft.

The FCC calls for all US-registered spacecraft to be disposed of at the end of its useful life. This means either decay into the atmosphere within a specific amount of time (25 years, I think) or placement into a "disposal" orbit. For geosynchronous spacecraft, that disposal orbit is one slightly higher, getting it out of the way of operational spacecraft.

Except that the method of government support in the past is completely different from what the FTC is suggesting here. Your own link talks about how support was given by publishing contracts to print laws, proclamations, etc, as well as reduced mailing costs and tax exemptions. What the FTC is talking about are direct bailouts and artificial restrictions on information (copyrighting facts). Seems to me that this is the antithesis of the spirit in which the early American government promoted the free press.

Proven approaches to what? Orbital fuel depots and refueling, inflatable aero shells, tens of kW electric propulsion for manned missions, and inflatable habitats don't have proven approaches yet. That's the whole point of these kinds of programs. The only way NASA will be able to stay with "proven approaches" is to remain in LEO and build carbon-copy ISSs. Even then, I remain far from convinced that what NASA does today should be considered a "proven approach" to manned spaceflight. While they've done some amazing things, human spaceflight still remains rare and hideously expensive. I would prefer that not be the approach to such matters going into the future.