In Italy they are discussing another hypothesis regarding the input power: The hypothesis is, that the load (the three inputs of the black box) were not connected between the phases and neutral, but between the phases. But, at the same time, the tensions were measured (by error or to mislead) between the phases and neutral.
In this case, the PCE830 computer "sees" the following: 3 times 6 A current 3 times about 230 V In reality, the power consumed was: 3 x 400 V x 6 A = 7200 W with pf=0.5 we have: 3600 W with 33.3% duty cycle we have 1200 continous power And that would be 3.3 times more then claimed, and would reduce the the so called COP of this nuclear reactor for domestic use to under one. Some commentator here believes to have seen on a picture, that the wires were in "trangolo" (triangle, between phases) and not "a stella" (star, between phase and neutral). What we know from the picture (that was take for shure DURING the afaik uninterrupted test), is that the tensions were measured between phase and neutral. Lets start to confute this hypothesis....
