[GMKlenklen]

Although i have never made a swervedrive, drivetrians are my "area of expertese" (although take that with a grain of salt, because i have only actually built one)
Assuming you can get any material that exists, i'd then start considering what you can do with it (what tools you have, how complex the cut outs can be, etc.) If you don't have a cnc mill, or a mill at all, then maybe some crazy aluminum contraption isn't the best for your team. Just ask the people who would be cutting/bending the meterial. lighter would be better, however, considering that a swerve drive is going to be a bit hevyer than kitbot drivetrains (generally more material and more complexity/parts)

There is a diference between independent (or paired) controll of modual rotation with respect to robot frame, and then haveing "one motor" like you said controlling all moduals, such that they all point the same way no matter what.
basically:
Independent controll means more motors, and slightly more freedom/complexity for the driver. also mechanically redundent as long as your wheels can be powered forward and back independently.
"one motor" controll means less motors, and generally makes things easyer on the driver. however to rotate the robot around an axis, two wheels will have to turn forwards, while two wheels turn backwards. lots of skiff here, because it's impossible to posititon wheels at a tangent to the "circle" your rotateing about. I hope you understand what i'm saying here.

Because i've never intended to make a swerve drive ( i prefer "holonomic" drives (mechanum or all omni-wheels)) i haven't the slightest clue how to mount a modual. I'd go find someone from "bomb squad" team #16. this year they sported a 3 wheeled swerve drive at the kansas city regional. they'll have some time to talk because I don't remember them winning anythng, although i don't know if they are goign to nat's just becasue.

hope that helps...