The idea of an extruded aluminum skin wrapped around an inflatable torus is an excellent idea. The torus can be sized to fit in whatever launch vehicle is available at launch time. The trick is to avoid having to launch EVERYTHING from Earth's surface.
The best way to decrease the cost of this exercise is to do this in a repeatable, reproducible fashion.
The single biggest cost is currently launch costs. The fastest way to decrease launch costs is to launch from someplace with lower gravity. The nearest place where we can find the resources for this type of exercise is the moon. The lunar surface is rich with the following resources: Silicon, Oxygen, Aluminum, Iron, Magnesium, Calcium, Titanium, Hydrogen, Helium, Argon, Nitrogen, Neon, Krypton, and Carbon.
So, what can we do with these resources? Well, first off, silicon, and various other materials can be used to create solar cells. Not necessarily high efficiency solar cells. But, if they can be manufactured in-place, on the moon, efficiency doesn't matter, because we simply trade # of cells for efficiency. Secondly, we can build capacitors. Big capacitors. And thirdly, we can build a rail-gun for launching refined materials back to earth-orbit. Fourth, we build solar powered smelters, capable of being used to refine aluminum, iron, magnesium, and titanium. Purified, and extruded as wire, they can be used on-orbit for a variety of purposes.
Additionally, Iron can be processed into high tensile strength steel, formed into functional length trusses, and also boosted to Earth orbit. While many of these processes may require human intervention, to set up, and get running, automation should allow for a limited manned presence on the moon, while a stream of prepared resources moves from the moon's surface to Earth Orbit.
By boosting the torus up, deflated, and then inflating it on orbit, it would be possible to wrap it with a combination of carbon fibers, and aluminum wire, to provide extremely high strength exterior armor. Once the torus is wrapped, trusses can be added, to provide mount points for all necessary fittings. The wrapping process could be highly automated, with relatively simple robotic wrappers slowly traversing the exterior of the torus, until sufficient layers of carbon fiber, and aluminum have been deposited. Alternatively, a layer of expanding foam could be laid down first, then carbon fiber, then another layer of foam, and a final layer of aluminum wire.
This would allow for some protection from micro-meteorite impacts.
Once the structure is completed, solar arrays (again from lunar resources) could be arrayed around the outer rim of the station, giving a sun-flower look, and leaving the central hub area free for maneuvering.
By having a two-part central hub, with one portion rotating with the station, and the remainder stationary for docking maneuvers, it should be possible to simplify the interface between.
Any zero gravity labs can be tethered nearby, with the understanding that access is by way of spacesuit. That way they will not be disturbed by any vibrations caused by the occupants moving within the station.
Finally, expansion comes in one of two ways. A second ring can be created (with a larger diameter) away from the current station, and flown into place when completed, providing a bulls-eye type arrangement, once attached, or a second torus can be inflated, and wrapped, adjacent to the first.
In either case, accessways would be necessary, but should be a relatively simple exercise, after all we've done so far.
The advantage with the "bulls-eye" arrangement is that you can have multiple levels of gravity on the station, for example, a lower gravity section, for those who would be unable to survive at full earth gravity.