You need to be careful here Jed. A sample of material is a finite object which can be tested as many times in as many different laboratories as its quantity allows.
Nope. Not in this case. The people at the NRL did all kinds of destructive analyses, stripping off layers and finally dissolving the samples. They told me, "the angels weep when they see what we did to those samples."
Heat as measured by calorimetry does go away but you can capture the data, and with a good cathode you can make it produce heat again and again. Heat as measured by some other techniques, such as melting things, leaves more permanent physical proof. Things seldom melt in cold fusion experiments, but in some of Fleischmann and Pons experiments the Kel-F plastic holder for the anode and cathode melted. You can estimate the highest heat from that. In many experiments, the water boils away. That would be an absence of evidence. As in Mizuno's grad student who asked, "Where'd the water go?!"
X-rays can be captured in a way that cannot be disputed or refuted, with x-ray film (autoradiographs).