Let's summarize a few things at this point:
1) RST predicts that different charging voltages will lead to different normalized charging speeds. By normalized, I mean expressed as a fraction of the charging voltage.
Experiment fails to show any such variation with charging voltage.
2) RST predicts that, at voltages well below about 5V, capacitors will charge much more slowly than predicted by normal theory.
Microprocessors, memory chips and other common electronics would fail to function if this were true. CAD tools used at companies like Intel would predict much higher possible CPU clock speeds than observed, and the industry would be in dire straits. No such catastrophe has occurred, and power supply voltages are around 1 volt, where the 5x predicted speed difference between RST and normal theory would obviously be too dramatic to ignore.
No such discrepancy has been observed. Electrical engineers are routinely able to predict -- quantitatively -- the behavior of real circuits, shipping in billion unit quantities annually, each chip containing hundreds of millions to billions of capacitors, to error bars within the tolerances of the components. That's a lot of experiments thumbing their noses at RST's absurd predictions.
3) RST's voltage-dependent behavior implies nonlinear behavior in capacitor circuits. Nonlinearity means distortion aplenty, and audio amplifiers would sound intolerable. Stereos are enjoyed by billions who seem not to have observed this predicted behavior.
RST -- and Transpower/Ronald Satz -- is simply wrong. Experiment, not opinion, has shown it.