Presumably hydrogen/deuterium, heat and/or electric current help out in some way, although how is not clear.
Yes, that is a challenge. But one good thing about the accelerated decay idea (or say the ULM neutrons) is that it narrows and defines the possibilities of what to look for from vast to modest. That is assuming the decay mode under stimulation remains within say the Firestone tables.
Robert Bryant's suggestion in this thread that the decay modes may be quite different certainly fits with the missing gammas for D+D by Pd electrolysis. But, combining that with the idea that "all modes of decay are accelerated" gets to be too much for the "narrowing" idea above.
There are enough preliminary "successes" in CF / LENR that a working device might be contrived for the purpose of examining a lot of possiblities in parallel (borrowing from what is called combinatorial chemistry, that is massively parallel manipulation of many variables in tiny reactions run simultaneously).... then going back and tweaking other variables or refining the combinations that worked best in prior iterations. I recall that Nikita Alexandrov once proposed such an approach at an ICCF talk.