The paper, titled "Observation of abundant heat production from a reactor device and of isotopic changes in the fuel", apparently follows the same thread of a previous publication; indeed the authors of the former and latter studies are partly the same (with the addition of Torbjorn Hartman). In particular, the technique devised to measure the radiated power of the reactor is the same, based on thermal imaging cameras.
The most significant addition from the former study is that this time the reactant (the substance contained in the device, which is subjected to heating by resistor coils inside the reactor and which allegedly produces excess heat) is studied for isotopic changes after a period of over 30 days of continuous operation. A further improvement is the careful monitoring of ionizing radiation emitted by the device, with the use of several independent detectors.
One further add-on is the analysis of the coating of the device, which might be suspected to be made of some special material. It is found to be aluminum oxide, so the explanation for where the excess power comes from must come from somewhere else. Also a direct-current component of the applied electrical power is monitored, finding it negligible. This probably is meant to address earlier criticism that some tricky wiring of the device allowed more power to be supplied to it from the ground wire.
Despite all the above important improvements, the study betrays one of its major shortcomings in clear already on page 7, when it clarifies that the inventor of the device took part in the experimental activities:
"Rossi intervened to switch off the dummy," (the reactor unloaded with reactant) " and in the following subsequent operations on the E-Cat: charge insertion, reactor startup, reactor shutdown and powder charge extraction."
This means that part of the test was conducted with the physical presence of Rossi in the critical phases. Here it is useful to remember one famous quote by Dick Feynman: "I believe a scientist dealing with non-scientific issues is as dumb as the next guy". With this I mean to say that the authors of the study, despite their curricula and scientific reputation, may have overlooked some simple trick. The absolutely mandatory requirement for an independent test is the total absence of the interested party from the scene. So it is a rather idle occupation to read the rest of the 53-page document.
For your benefit I did read it anyway, and I can summarize here the main points: first of all, the heat output measured by thermal cameras has been calibrated with a dummy unit (a reactor with no reactant inside) subjected to some input power (a smaller one than the power the full reactor was subjected to, as the researchers feared to damage the coils ?!). The main measurement of the working E-cat has then been performed at two different operating points for 32 days. The result is that the output power far exceeds (by a factor of 3.5) the input, and the difference cannot be ascribed to chemical reactions - the mass of the reactant being too small to produce similar yields. The total produced power in 32 days of operation is claimed to be of 5.8 MJoules, with a production of 2 million watts per kilogram of reactant.
A second point is that no radiation has been detected to come out of the unit. So we are asked to believe that not only nuclear reactions have happened inside the reactor during operation, with no known mechanism allowing such a thing, but also that those mysterious nuclear transitions involved only emission of heat and no x- and gamma rays, or neutrons.
Then, the analysis of the isotopic content of a few grains of the reactant is rather sparsely described in the paper, but an appendix discusses it in great detail. The summary of the scanning electron microscopy analysis is that the "ash" collected from the reacted material has an inverted population of Li and Ni isotopes - with Li7 abundance dropping from 92% to 8% (with Li6 passing from 8% to 92%) and the Ni62 abundance raising from 4% to 99%. This result is indeed quite startling, and appears confirmed by an independent analysis performed with a different technique based on mass spectroscopy.
All in all, the study seems to overcome many of the shortcomings of the previous publication, except for the fact that the reactant was inserted and extracted by Rossi himself! Given the extraordinary nature of the claim to have produced a cheap way to extract energy from ordinary materials, through a unknown physical process, to me this constitutes a major flaw, which totally invalidates any conclusions one might otherwise draw from the data alone. It would be like if I asked you to believe that by putting a dollar bill in a special laundry machine and spinning it for half an hour with some special detergent the dollar turns into a $1000 note. You are allowed to watch the machine as it does its work, but it is me who opens it and extracts the bill when it has finished its magic conversion. I doubt you would buy it.
Alas, I feel the problems of mankind will not be solved by magical tricks. Hard science is hard not only because it is difficult to study it and further our understanding of nature; it is hard also because it yields no free lunch.
UPDATE: a much more complete view at the whole issue is here, with links to a lot of material that readers of this post need to give a look at, if they believe the E-Cat could really be a working thing.