Seems to me, in addition to lighting, a
steady/still camera and good composition, that the quality of the camera
itself must have a lot to do with it.
When Coop showed us how to make one of these at Wulf's hammer in in VT
last summer he used all of his top end SLR (with his professional lightbox), his
pocket camera (with the cheap lightbox) and a random camera from a woman
in the audience (with the cheap lightbox). You could tell the difference all
the way down the line, but even the last produced better results than most
of us are used to.
He hand held all the cameras. For the SLR (Canon 5D as I recall) he uses
50mm 1.4 and 100mm fix focal length lenses. Says that they give better
results than zooms for this type of picture. They're cheaper too.
Actually, the only really expensive item in his full professional setup was the
camera body. The box was a bigger piece of artists velum stretched over a
PVC pipe frame. The shadow reflectors were mylar and plastic (e.g. a sheet
of reflective mylar on a plastic cookbook holder) and movable. The lights
looked exotic, but they were a 20-30 year old professional flash rig that he
said you can get on eBay for around $300.