Unfortunately you have a fundamental misunderstanding of health experimental design. Double-blind means that neither the participants nor the experimenters know which participant is allocated to which group. It doesn't say anything about how many groups there are or what is done to each group. I haven't heard of Dr Bredesen, but they could certainly run a double-blind study to validate their treatment regime. Treatments that haven't been validated in this way have a high risk of being ineffective but appearing effective due to experimenter error, placebo effect, demand characteristics, etc. Requiring convincing evidence in the form of a double-blind trial isn't purism, it's sensible caution based on experience.

edit: you're also on shaky ground statistically. There's a rule of thumb that if interactions between several variables are present in a treatment effect, 70% of that effect will show up on the individual variables (as main effects). It's very unlikely that there exists a treatment where all the effect is on a 35-variable interaction and none of the effect is visible when looking at each variable singly.