Although I'm not a very regular user of modeller, I believe it is well known that long gaps, especially terminal gaps, look strange when modeled using default options. I was wondering, would be possible to add an Rg restraint so that such regions at least stay nearby the rest of the (well-modeled) protein? This is not just a problem of aesthetics. When I try to superimpose such structures using structural alignment, sometimes the long gappy regions throw the superpositioning calculation off. I could just write a script to try to cut them out, but it would be nice if I could fix the problem at the source. My suggestion of using an Rg restrain would be basically 

"if distance of residue x from center is greater than some huge value (2-3x the predicted Rg, based on sequence length), add a bias to the potential for x toward the center of the protein"

This kind of restraint should leave the "good" parts of the alone and just pull the "bad" parts closer to the rest of the protein.