The ridge of the duocylinder (which is related to the clifford torus) is a uniformly-deformed torus. It's not quite flat, because there is still deformation, but it's different from a 3D torus where deformation is non-uniform. However, I don't think the duocylinder ridge is what they're talking about here. I'm not sure I understand how the smooth fractal makes it non-deformed. If anything, I'd say the duocylinder's ridge is closer to being a "faithful representation" of the flat torus. But I may be biased.