I don't think it is "disturbing" that CEC declines as process temperatures rise, but rather understandable. Again, the higher the temperature, the longer the residence time, and the more oxygen that is introduced into the pyrolysis zone, the more oxygen, hydrogen and carbon atoms that are not bound in aromatic rings are driven out of the char. Those are the elements needed to form negatively charged functional groups on the surface of the char. The more the evolving char becomes a regularly ordered structure dominated by aromatic rings - or the more it is altered from its original, somewhat chaotic structure, the less it is able to enhance CEC.
It could be that some data shows an increase, say between 400 C and 500 C, but looking at the molecular structure of humic substances and how they enhance CEC indicates that is is very likely we want to preserve a similar molecular structure in biochar if we want it to have the same effect. Data showing a decline of CEC as process temperatures rise seems to confirm this line of thinking.
See attached preso that I developed. The animations are missing in the conversion to PDF, but the only detail that isn't well conveyed is in the slide showing the root and root hairs. A carbohydrate molecule is shown, but what isn't clear is that the plant barters a carbohydrate to the via the hyphae of mycorrhizal fungi in exchange for a needed cation or anion.
I'm not a soil scientists, but the ones I work with here have reviewed the model in the preso and have told me it represents how cation exchange is understood to work.