A) Did you really mean Rosetta 3.0? I strongly suggest upgrading to 3.2.
That paper was done in Rosetta++ (one of the 2.x versions). The precise protocol was never ported from 2.x to 3.x. I’ve asked the authors what they think about getting it done in either version.
C) The hardest part is likely to be the fragments. The easy way to generate fragments is with the robetta server (http://robetta.bakerlab.org/), but it only makes 3mers and 9mers. You’ll need to learn to make your own fragments to get 10mers. The old-style fragment generating code is included with 3.2; there is also a powerful fragment picker included inside 3.2. Unfortunately I know how to use neither of these.
D) The existing loopmodeling executeable is not set up to perform design, but that can be changed in a relatively straightforward fashion. It may be easier for you to write your own script (using the RosettaScripts tools) than change loop modeling, it will be your choice.