Mike Here is why you are puzzled.
You say: "At some point the NaH decomposes, releasing Na and H atoms in close proximity, whereby Na++ then catalyses the H producing H[1/3]. There are aspects of this which puzzle me." No Kidding! Not the least of which puzzlement should be that this species is NOT a decent fit for a catalyst under Mills' CQM criteria. I have a version of CQM written in 2001. Now you may want to say that Mills has "refined" things since then, but I say instead that he has "shoehorned" them, based on a lucky finding of an energy anomaly with sodium. For the moment - let's say that back then after nearly a dozen year of going at it - he should have been able to tell what was, and what was not a catalyst, under his theory and here is what he where he places sodium on page 147: Na+ the ion, and not the atom - becomes a catalyst - only when - forced all the way to IP4 by adding the enormous energy of almost 218 eV per atom (a fairly strong x-ray) which will never happen, even of on the far end of Boltzmann's tail - when the input to his reactor is considered. Basically he rejects the idea that Na++ is a catalyst and says that not only must you start with the single Na+ ion, which is no problem, but then all at once you must remove AT THE SAME TIME three additional electrons and with a proton in the vicinity. This is unrealistic, of course, in that type of reactor. This is why I claim that he is shoehorning "lucky" results, found in experiment, into the theory when in actuality - it is very likely that something else is happening. Jones