That is exactly what we plan to talk about. The calorimeter becomes part of the system and they influence each other. The calorimetry can be perfect but if it removes too much heat in the measurement process, then excess heat will be difficult to detect.
Yes. This is why I have been developing a simplified model of a heat-activated exothermic process in a thermal mass. The model is robust. It depends on the topological relationship between the temperature-dependence of the LENR activation and reactor cooling curves rather than on precise parameter values. The model predicts that sort of heating that Daniel_G claims to have should result in inflection points in heating and cooling curves, hysteresis with respect to the input power, and temperature escape to meltdown. It wil be interesting to see if any of these things show up in Daniel_G's data.