This is actually incorrect; it is, indeed, a loop of rope. See here (
http://www.conterra-inc.com/images/deta ... bready.jpg) and here (
http://www.conterra-inc.com/images/deta ... abbig1.jpg).
It's exactly the same situation you see when you run a rope over a hyperbar, as Scott explained. Using multiple "horns," as shown here (
http://www.conterra-inc.com/images/deta ... abwrap.jpg), would appear to twist the rope horribly. Likewise, this double-rope method involving two horns looks like a mess waiting to happen (
http://www.conterra-inc.com/images/deta ... arabII.jpg).
This thing reminds me of a big, stainless steel version of the Petzl Pirana, the primary difference being the large tube-shaped wear surface in the center; where the Pirana is based on an 8, this appears to be based on a rack or a brake-tube style descender. While stainless steel descenders can certainly be more viable for caving than aluminum ones like the Pirana, adding a half-dozen "additional friction" options doesn't necessarily sidestep certain limitations. A rack with eight hyperbars is still, fundamentally, a rack.
Braking carabiners used in conjuntion with a bobbin descender are a different animal; a standard "carabiner next to bobbin" approach won't twist the rope, but I suspect using a Freino might twist it a bit.