B. Has this line of research actually shown anything in a bulk setting at sane levels of incident RF or is it still "we built an antenna structure out of saline and showed that a neuron can respond to the resulting voltage if we really blast it?"
Thermalization mechanisms like rovibrational absorption (small scale) and conventional currents (large scale -- the focus of these new studies) aren't the only supporting evidence for thermal damage models, though. There is also the dose/response curve. We can generate RF at many thousands of times the intensity that we find in cell phones and observe the consequences. Often unintentionally. Those consequences look a heck of a lot like thermal burns. If there is an alternative damage mechanism, it has to saturate at a very low level, so that there are no observable acute effects from a wild overdose.