A method for quick RF electrolysis could be useful. Perhaps it requires smaller electrodes, fewer electrode replacements, or less overvoltage (inefficiency) than DC electrolysis. Such a finding would have very real commercial implications, assuming it isn't common practice in commercial installations already, though if it is, I haven't heard.
Unfortunately, he also claims to have found free energy:
> The flame generated by the device is useful both for heating various substances and as a source of clean energy, capable to substitute all types of fossil fuels
Turning H2O into H2 and O2 is the reverse reaction of burning H2 and O2 into H2O. If this roundtrip is done with perfect efficiency, the device will not consume any energy, but it will also not produce any energy. In practice, it would be extremely difficult to capture 100% of the heat and light emitted by the flame, so the device will merely consume energy. It's useless as a free energy device.
> causes NaCl solutions in water, with concentrations from 1 to over 30%, to be measurably changed in structure
Null hypothesis: the "changes in structure" are the exact same you would measure in any other NaCl solution with increasing concentration. At the very least there should be a control.
The PDF paper, which is the only peer-reviewed part of this whole mess, makes no claims of free energy and fails to investigate the obvious null hypothesis for their observed effects.
Absurd perpetual motion claims and diversionary "changes in structure" nonsense make me doubt this guy has the chops or inclination to pursue commercial applications of RF electrolysis, even if his findings are real and have commercial applicability in that direction. Verdict: