So that "plate" section on its' edge is the only factory roll resistance at the rear?[/b]
Do you mean the vertical part of the beam? If so, yes. Those "plates" that attach the two trailing arms together are both structural items to keep the arms in place *and* torsion bars to resist twist. Anything you can do to increase the "torsion" of those plates will increase roll resistance in the same manner as an "anti-swaybar".

This blue bar bolts to that plate and adds to its' rigidity?[/b]
Don't get me wrong: yes it does. But, since it's not acting co-axially to the axis of bending of the trailing arm assembly, it's acting more in tension than in torsion (though, you MEs will argue with me that a torsion bar does, too. I hear ya).

Are there clearance issues that would prevent a more common bar design? Why not add "lever arms" on this bar that attach directly to each trailing arm?[/b]
Haven't been under the Professor's car in a while - most of my experience was in Rabbits - so I'll look next time I'm there. What I was basically trying to indicate is not that this bar is ineffective - it is - but it's not *really* doing the same thing that a torsion-based anti-swaybar does. This bar is acting more as a stiffener.

Another way you could re-create this action - as a "for example" - would be to use flat plate and make those triangular tabs one big plate across. You could then see that as the beam is twisted the plate would be "stretched". Then you could add a vertical angle on the front edge to increase its torsion, and you'd be effectively doing the same thing.

The Professor and I have discussed this, but given the whole rear suspension is one big anti-swaybar design, and rear swaybar are free, I think there's a LOT of clever things you can do with this that are well within the letter of the rules. However, I don't see the rear suspension design of this car as the limiting factor; you'd get a lot more performance out of trying to make the Mac front end work better...