Fair point. I was using "controlled" as a shorthand for the usual qualities (double blind, randomized, sufficiently many test subjects, etc) you'd associate wit

Fair point. I was using "controlled" as a shorthand for the usual qualities (double blind, randomized, sufficiently many test subjects, etc) you'd associate with good experimental design.