It appears that most of your comments have been directed to Evan Liu. Since the three organizers played an equal role in organizing the competition, Kevin and I (Mitchell Stern) have decided to make the following posts.
We would like to formally apologize for the organizational issues at our...
Werner, this topic is about swapping adjacent cross edges without disturbing the already-solved F2L pairs.
For swapping adjacent cross edges right after making the cross, I'd suggest a slightly shorter alternative: z' U' R U R' U' z.
To swap DF and DL, I usually use M' y' M' U M y U2 M. It's 100% intuitive and fairly fast as well.
To swap DF and DR, you could either do a rotation first then use the algorithm above, or use the mirror, M' y M' U' M y' U2 M.
To clarify:
1) Is the wall infinitely wide as well, or does it have a finite width?
2) Is the mass density of the wall uniform, or is it distributed in such a way that the wall has finite mass?
If you use the Roux method for a 3x3, why would you use reduction for big cubes? Methods like http://www.youtube.com/watch?v=llhAX09mjIk or K4 are surely more suitable...