The article intro above actually explains this, if you read it. The fuel in this tank is BURNED, getting the payload into orbit. In the Apollo mission days, the payload was e.g. a third stage that went to the moon and back, as these are BIG rockets. In the past, the second stage tanks would be "thrown away" and allowed to reenter and burn up, but that's slightly insane given the roughly 32 MJ/kg direct energy cost (multiplied by a few orders of magnitude) of lifting anything at all into orbit.
The reasoning is then as follows: We've gotten this great big cylindrical chunk of pressure-tested metal -- remember, it held liquid hydrogen at HIGH pressure securely through a launch exerting many g's of acceleration -- into orbit. It already cost us millions of dollars to build, and tens of millions to get it into orbit as a SIDE EFFECT of lifting this other, really big payload. Let's not waste it!
So, what can we do with it? Well, given that it is roughly the size and even the shape of a good sized mobile home or the living volume of early submarines, making it into pressurized living space is an obvious choice. It is pressure tested at many times the 0.5-1.0 atm pressure differential needed to sustain human life in space. It is made of high quality, carefully x-rayed, stress-tested metal (because NASA would be insane to fire a rocket into space with humans on board with anything less holding in the fuel of the rocket). The metal has been carefully crafted and annealed to be able to handle liquid hydrogen temperatures without becoming brittle, so it is also proofed against your concerns with heat -- humans cannot tolerate any temperatures this metal is unlikely to be perfectly capable of withstanding, and besides, shielding it from sunlight is a matter of wrapping it in a reflective mylar blanket that weighs almost nothing and can easily be shipped up as part of the conversion kit.
As for radiation shielding -- that I don't know about, but I very much doubt that it is an issue. If the Earth gets hit dead on with a solar flare, I don't think there is anything we could reasonably put humans inside in orbit that would be "safe". It's not clear that being on the Earth's surface inside the atmosphere would be "safe". If the metal that the container was made of wasn't adequate as shielding during such an event -- I'm pretty sure it would be perfectly good most of the time -- and we had something better (but smaller and more expensive) then humans could retreat into the latter as a "shelter" to wait out the storm.
Life support machinery and furniture for the interior of the tank turned into habitat is a small fraction of the weight of the whole thing, and weight into orbit costs like gold.
Now let's compare costs. Suppose you used the Atlas to launch an Earth-built space habitat directly into space as to you suggest, and just wasted the second stage tank as usual. It costs you one launch to get the habitat into space, and the interior volume is almost certainly going to be smaller than the second stage tank volume. Now suppose that you take the empty tank and just hook it onto the habitat you just launched (which already has all of the life support machinery, radiation tolerance etc that you are worried about. Voila! You've more than doubled your available habitat volume in space at (almost) zero additional marginal cost! EVEN if it isn't AS safe as the primary habitat in the event of a solar storm, well, astronauts can always retreat into the primary habitat during such a storm and still use the tank as room for experiments, hydroponics, their ping pong table, room to spread out in to avoid going nuts.
The last question is: What do you have to do to the tank to FACILITATE this so that it isn't being done on an ad hoc basis? As you say, certain pieces of work are way cheaper on Earth than they will be in orbit. Should we build the tank out of slightly different metals so it IS a better radiation shield? Should we pre-install ductwork for ventilation and wiring and liquid management (water and sewage and hydroponics) while it is still down here on Earth? Will any of this compromise its function as a fuel tank (almost certainly not, up to SOME point, but what is that point)? Can we design it so that converting it in orbit is a matter of removing this big plate here, mating a now-exposed sealing ring to a matching ring on a modular component, hooking up "life support" in a daisy chain or fastening it onto the outside and connecting it up, and then just moving in? Remember, this is where human creative design can really shine -- maybe we can ship the furniture and hardware needed to complete a standard conversion INSIDE the tank or INSIDE the otherwise wasted interior volume of the primary payload. That's the whole point of the study. Who knows, maybe they WILL conclude that it isn't cost-efficient or feasible, as this is a feasibility study and if we knew the answer, why bother with the study? But the reason for the investment is clear -- $65 million is chicken feed compared to the potential cost benefit of saving a SINGLE launch of a SINGLE Earth-built habitat into space.
No matter how you slice it, using the tank very likely nearly doubles the USABLE mass you've just lifted into orbit, and if we cleverly re-engineer the tank to be BOTH a tank AND a habitat shell, we can probably reduce the third stage "payload" to be nothing more than the pre-fabricated, modular support hardware. It is almost certain, note well, that this is going to be the MOST cost-efficient way to get ANY pressurizable living volume into space. In a possibly mythical Mars mission, maybe we'll end up using it again first as a -- fuel tank -- refuelled from a comet head we've grabbed and towed in and kept on ice for that purpose. Then as the tanks empty along the way, they become living space for the crew. Maybe they can be daisy chained together to make an actual space habitat in the form or a large rotating ring with "gravity", a design that dates back to early science fiction, after being towed (say) to a Lagrange point using a solar sail (free transport, once you make it into orbit, if you're not in a hurry). The point is that worked metal in space is worth its weight in gold (or in any event costs almost its weight in gold to get it there. Don't waste it.
To quote Robert A. Heinlein (loosely) "Once you are in orbit, you're halfway to anywhere" -- which energetically is precisely true.