That M has not experienced meltdown or "saturation" in R20 is a bit surprising and rather goes against the other things he has observed.
Something that has not yet been mentioned here is that absorption of Deuterium by metal lattices is an exothermic process, while releasing of Deuterium by the Pd/Ni lattices is a endothermic process.
I don't have numbers on this process available, but this 'breathing' of Deuterium in a kind of semi-equilibrium may work as a dampening effect towards runaways, in particular where high absorption is never reached.