NONE of those explain the change in isotope species described in the article. Unless you mean he is somehow beaming concentrated neutrons through some unknown means into the device somehow, or that he is able to somehow completely replace the device with progressively more concentrated populations of heavier isotopes miraculously every time the researchers check.
Occam's razor sometimes shows that the seemingly improbable is actually the most likely explanation.
The actual definition of that particular by-rule, is that the explanation with the least complications (read, elaborate conjectures and weasel wording) is the most likely to be correct.
This device appears to produce power with an energy density many times greater than dynamite, and produces a change in isotope species of the test sample.
At this point, the test for fraud is to determine if the calculated energy released is congruent with the change in the mass energy potential of the sample before and after the experiment. Conservation of energy says that if this device used fusion, or any other nuclear power based reaction to achieve its outcome, then that energy came from mass.
They measured the energy released. Measure the difference in the mass of the sample after the 32day observation period, and compare it against the mass of the sample before the observation period. If the calculated mass value for the energy released + current mass == mass of sample before observation, we have a very difficult thing to account for, because it means the device is plausible.
If they dont match, it means the man is a fraud.
This is a testable point of data that would make fraud detection very easy, and would make people that are quick to point the fraud finger very uncomfortable if found to be true.
If the researchers did not collect this measurement, KNOWING that this device was 'supposed' to produce power via a nuclear energy process, then you have a very good grounds to seriously torpedo their published paper, and recommend additional experiment due to improper testing process. Especially since they have the equipment to measure statistical isotope species in the sample, and knew to test for it.
Granted, the difference in weight for a 140mwh value would be in the picograms or less. That just means that smaller samples with the same reaction process need to be studied so that tinier and more sensitive aparatus can measure any changes-- which would also make the "he switched the samples!" argument more convoluted in such latter experiments.
Of course, the NOT SCIENTIFIC AT ALL approach is to just say "There is no need to conduct that experiment, because it is clearly a fraud!"-- That's the not-science-at-all version of begging the question in a wrapper of appeal to authority fallacy, DRESSED as science.
Science is about observation, and recording data about observation, and making hypotheses that predict future observation. In science, REALITY IS KING. If the experiment has shown that energy was generated, and specific features were measured, but the experiment itself is in question-- the proper course of action is to repeat the experimental protocol in additional laboratories to eliminate the conjectured disqualifying uncontrolled variables cited.
In this case:
Were there any spurrious or anomalous EM readings near the device? (Any "beamed" energy delivered to the device would have to be of this type to interact reliably with electrical energy metering equipment.)
Was the sample ever tampered with? (Repeating the protocol simultaneously in multiple labs around the world to verify the results would exclude this, discounting some brilliantly absurd conspiracy.)
Were the researchers involved in any lucrative scam tactics with the inventor? (again, more independent testing would reveal this.)
So, in all cased, the prescribed course of action to verify definitively that this device is a fraud IS TO DO THE DAMN EXPERIMENT, REPEATEDLY.
THAT IS EXACTLY WHAT KILLED THE ORIGINAL "COLD FUSION" DEVICE. IT DID NOT WORK WHEN REPRODUCED IN OTHER LABS.
These findings suggest that further experiment is required.
It does not definitively prove the device works as advertised, and it does not constitute proof of fraud. It proves that further experimentation is justified, and necessary.
So, for all those morons who are shouting that this is a fraud straight up without doing the experiment, I want to see your data. No, Not other people's data. I want to see YOURS. SPECIFICALLY.
What's that? No data? Then fuck off. You don't further science by NOT doing experiments. That's the purview of philosophers, politicians, and religious dogmatists. It is PRECISELY WHY SCIENCE IS DIFFERENT.