There's no clinical trial you could possibly do to show that these machines are "safe" in the medical sense (i.e., that the benefits to human subjects outweigh

There's no clinical trial you could possibly do to show that these machines are "safe" in the medical sense (i.e., that the benefits to human subjects outweigh the harm). You could expose animals to the same doses of radiation and see what happens. This has been done for the general types of radiation (with X-rays animals contract cancers and die at sufficiently high power, with millimeter waves there's substantially less evidence for damage), but AFAIK not with the machines themselves. The biggest danger may not come from chronic exposure, but from an acute failure of the machine or mistake made by the operator, the probability of which could not be determined in a study anyway. (Therac-25 delivered overdoses of radiation to several patients due to a software bug; Cedars-Sinai recently reported that they had accidentally delivered dozens of CT patients doses several times higher than normal; patients get killed by oxygen tanks and other objects sucked into the magnetic fields of MRIs with startling frequency.) In any case, I think the danger posed by millimeter wave systems is significantly less than that of backscatter machines. There are plenty of studies to show that microwaves don't cause cancer, and it's hard to imagine a malfunctioning machine delivering something harmful.