Someone for the love of God please explain to me why replicate an experiment that makes "40 to 100W" while the newest claim from the same experimenter is a reactor that sits quietly in his fireplace, making 3kW from a 300W input?
If people skilled in the art trying to replicate get ~10 W the first time around, it will be a triumph. It will prove that Mizuno is right.
If they keep at it for a year or two, I expect they will get kilowatts for that mass of reactant (~54 g). More likely they will want to work with a smaller mass of reactant at lower power, at a similar temperature and power density. It is much easier to work with 10 to 100 W than kilowatts. It is easier to measure that accurately. That reactor is so big, it is unwieldy. It is big because Mizuno thinks a large mass of reactant is more likely to work, and because if it only works at a small power density he can still detect it. Last year he was getting ~12 W out of that much reactant, which was close to the margin to measure.
The kilowatt scale reactions are probably mainly due to the improved design. That is the only design we described and illustrated in the paper. So that is the only one people will try to replicate. That does not mean it will work. Even if it does work, that does not mean it will work as well as this reactor did. That would be a miracle, not a replication. That never happens the first time around in experimental science. There may be large variations in Pd-on-Ni reactant performance. The high power may be partly a matter of luck. Mizuno himself might not see it with the next batch of reactants. That is hard to predict. After a few years and several hundred million dollars of R&D, people will probably figure out how to make the reactant more consistently. A few years after that, and a few billion dollars later, they will have far better reactants and better control. I suppose the reactors will look quite different as well.