In light of Rossi's apparent lead i'd be looking at the possibility of spontaneous formation of novel condensates. The D2 diatomic molecule being a boson presents an obvious soft target for aligning spins to cohere into shared lower-energy quantum states, the different magnetic moments of the electron and nucleus of the deuterium atom passing through the high magnetic moments of the Ni powder might cause some degree of polarisation and/or phase coherence, etc. - the basic idea being to cultivate an optimal fluctuation in Fermi numbers / system entropy relative to surrounding environmental free electrons (such as in the reactor casing, say), which may adopt the lower-energy quantum states vacated by the formation of the condensates; thus forcing the vacuum to assign higher Fermi numbers / quantum energy states to any fermions exiting a collapsing condensate than they initially carried into it. This manifests as an exothermic gain accumulating over many such cycles..
..until thermal phase begins to approach resonant modes with certain quantum phases (such as the zitterbewegung phase), resulting in further quantum-classical coherence and allowing like-polarised electrons to begin sharing Fermi numbers, so adopting the lowest available energy state but also causing precipitous fluctuations in system entropy relative to the environment, surrounding free electrons instantly co-opting lower Fermi numbers as they become available, and so causing the vacuum to assign necessarily-higher quantum energy states to fermions exiting transiently-coherent quasi-bosonic states and yadda yadda runaway feedback loop. Vacuum / ZP energy, bashically, a la EM OU - the accelerated / heated fermions exiting the meta-stable bosonic states being impelled by a flux of positive h-bar endowed by virtual photons corralled from vacuum per QED - ie. the actual form of the gain being normal Coulomb repulsion between decohering fermions, albeit with vacuum-inflated quantum energy states. LENR effects may be epiphenomenal to the common energy gain principle, an almost incidental artefact of spontaneous long-range coupling between nuclei immersed in a matrix of coherent quasi-bosons allowing them to interact in some sense as if already within one another's proton radii, presumably conserving baryon number if not initial disposition (again, a different kind of effective Coulomb exploit). The key dynamic would simply be that transiently-stable shared lower-energy quantum states free up Fermi numbers that any environmental fermions in higher energy states will automatically drop down into, necessitating the assigning of higher Fermi numbers to those departing these shared states, thus energising these collapses with an extra kick of ambient h-bar from vacuum. This is basically what Rossi's doing via contained cold plasmas - time-critical selective-energisation of electrons over ions causing like-polasrised condensates of the former to precipitate out onto the latter, forming transiently-stable structures with an emission line consistent with the precessional moment of a protium-nucleated thermo-ZBW condensate; decoherence of which (reinstating Pauli exclusion) yields excess energy, the cycle requiring cool-off time to prevent ions gaining thermal equilibrium with electrons (via normal dissipation), hence a discontinuous / pulsed cycle, in order to maintain the formative 'non-thermal plasma' state for consistently culturing these exotic self-reactive species and the huge fluctuations in internal vs ambient entropy / quantum energy states their concerted phase transitions apply. Crucially, the system remains thermodynamically open to (and dependent upon) the environmental exchange of entropy, 2LoT itself putting the vacuum in a bind, which then HAS to assign higher energy states to decohering fermions due to their formerly-vacated states having been reoccupied the instant any up-up or down-down condensate initially formed.. and so a 'logic trap for nature', type stuff. TL;DR - mebe the Thermacore runaway was inadvertent EM OU resulting from exothermic closed-cycling of spontaneous fermion-boson-fermion phase transitions - perhaps helped along by chance resonance with ambient phonon phases - between molecular and atomic D2 - until reaching breakdown temps, at which point the fermionic D2 atoms become bosonic ions and fermionic up/down electrons which then want - through shear thermodynamic expediency - to form like-polarised quasi-bosons sharing a unitary quantum energy state; everything's jumping back and forth across the Pauli exclusion barrier in sweeping phase transitions - that's just what happens when a gas of atomic and molecular D2 gets ionised - and because nature only cares about balancing the books in the given instant, refermionisation events per se (such as bosenovas) may be ideal bait'n'switch / shell games for wrangling unilateral Coulomb repulsion from ZPE. Rather than a 2LoT violation, a kind of negentropic open system.. Compounding anecdote with wild speculation, no doubt, but a different angle no less..
