To avoid these discussions about thermal inertia, it would have been better to log only end temperatures where the temperature is stabilized in time at a given input power. The same can be done with the active run. Calculate a curve through the calibrated points (temp. Vs power) and use this equation to calculate the power at the logged active points temperatures and subtract that from the input power in the active points. The result are excess power points that can be plotted against input power or temperature. It is a bit more work, but a lot more clear to show.
Anyway I think this whole debate is spurious as is only supported by the denial of the possibility of LENR being real.
Would we present this as the results of burning charcoal inside the same reactor no one would doubt the calibration vs active excess heat that would come from the charcoal.
If the only difference between the calibration run and the active run is the presence of the treated mesh inside, even if considering that the calibration run has not any mesh inside (I really don't know, but I say that even if that´s the case), any conventional explanation for that amount of excess heat can be safely ruled out. That leaves only unconventional possibilities.