I am pretty sure that all researchers want to keep their ingredients so extremely pure that they thereby eliminate the elements that are essential for the LENR reaction, and I am thinking primarily of carbon.
If the reaction is occurring predominantly on the outer surface of the mesh, carbon could possibly have a role and there are instances where it might be inadvertently indirectly introduced in a less than optimally maintained high vacuum system. As far as I am aware of, it's important to have a slight surface layer of carbon on the catalysts in Leif Holmlid's experiments; in his earlier works this often came over days/weeks of activity mainly from diffusion pump oils volatilizing in his ultra-high vacuum chamber and decomposing on the catalysts. I believe its main effect here is through advantageous modification of the catalysts' work function, but its precise role hasn't been detailed yet in a paper from his group. I can provide some references about it if interested.
If it's instead occurring predominantly at the interface between Pd and Ni like Edmund Storms predicts, then the surface state (e.g. oxidation or presence of other impurities) of the mesh/substrate when the burnishing step is performed could be important. In his paper on the process that I already linked earlier he suggested:
[...] The second requirement involves the small particles of NiO that would be removed from the Ni surface and mixed with the Pd layer. A gap would be expected to form between the surrounding Pd metal and this inert inclusion, as hydrogen is lost from the structure.
If this description were correct, it would be expected to apply to all Pd found to produce LENR. I predict that commercial Pd observed to support LENR contains similar unintended inclusions that remain in the metal after the refining process and were not altered when the metal was formed into wire or sheet. Consequently, most pieces of the batch are found to produce LENR regardless of the final form created by physical means. This behavior explains why some batches of commercial Pd produce LENR for no obvious reason.
As an example, Fleischmann has described how boron is added to the molten Pd during the purification process to remove oxygen by formation of insoluble boron oxide, which floats to the surface and is physically removed. In view of the Storms model, the small pieces of oxide scattered throughout would make the Pd eventually nuclear active rather than the absence of oxygen.
So, also according to this model, an excessively clean material may not work well or possibly not work at all in many cases.