One more 'flash from the past' on Robin's mention of an energy "sink" being "like a hole in the ground." (how quickly we forget)
This might also serve as some insight wrt Ed's comment. There is a geometric dimension to a 27.2 eV 'hole' if one wishes to consider the wavelength of UV photon radiation at this level. Methinks it is about 46 nm - which is well within the capability of micro-lithography at chip labs these days. Could it be that the Raney nickel being used was chosen, inadvertently or specifically - to have an average pore size near this dimension (46 nm) ? If so, then this makes a lot more sense as a "package" or as a system which can be scaled up - as this gives you the first level of 'shrinkage' very "cheaply" so to speak ... simply apply a positive charge to a geometric hole instead of 'manufacturing' one. And it could well be that after the first redundant level has been reached without recourse to adding massive amounts of energy, that everything works much smoother thereafter. The really interesting thing, from the perspective of LENR is that if it a geometric "hole" of this size (46 nm) works for protium, then it should also work for deuterium and could possibly increase the reaction rate when there is Pd in there instead of NaH. Why? Well for one thing - since the atomic volume of the deuteron is reduced by a factor of 8 (the cube of halving the diameter) then as much as 8 times more deuterium should fit into a Pd matrix (than normal) and there is evidence that higher loading is more active. This might also favor titanium instead of Pd as the active matrix - since its lower loading level would not be the limiting factor it is now, and since it is much cheaper. Plus - it is also possible that the Raney metal alone will be very active with deuterium. Mills may have opened a Pandora's box of LENR sectrets. Jones