Let me enlarge upon the difficulties you face in applying results of the mainstream electron screening experiments to PdD LENR. Assume 1 W excess power produced from a small palladium electrode. Now assume that the power is produced by fusing deuterium in a d(d,*)4He reaction, where * fills in for the missing gamma photon. 1 W power is 1 joule of energy per second. To get 1 joule of energy from d+d reactions, we must have:
1 Joule * 1 reaction / 24 MeV * 6.24e12 MeV / Joule = 2.6e11 reactions (per second).
What must be the efficiency of the bias of the branching towards our preferred branch of dd → 4He in order not to detect neutrons in a neutron detector above background or make the inventor sick? Suppose somehow we have an excellent efficiency of 99.9 percent of all reactions yielding the 4He/not-gamma-photon branch. That leaves 0.1 percent of reactions that produce the other branches:
(0.1 / 100) * 2.6e11 reactions / s = 2.6e8 reactions / s
producing something other than 4He and not-gammas. In other words, on the order of 1e8, or 100,000,000, neutrons per second. It seems our efficiency of 99.9 percent of good reactions is far too low. Whatever is using deuterium fusion to produce LENR along the lines of the mainstream electron screening experiments must be orders of magnitude more efficient at avoiding the usual branches in order to match the results of LENR experiments. Our LENR version of dd fusion must be nearly perfectly efficient.
The difficulty, of course, is that the mainstream experiments give no hope of such efficiency, no matter how much screening occurs. This should at least provide motivation for using lateral thinking to come up with another plausible explanation.
All the more reason to do some due diligence regarding whether the common understanding of deuterium fusion is a good fit for the facts. Perhaps there are other explanations that do not require such narrow tolerances and so many ad hoc assumptions.