In my experience, trying to assess the effects of a medication or technique on yourself is an exercise in frustration. Although you can do legitimate double-blind experiments on yourself, it's usually impossible to properly control the variables which can affect the outcome (diet, exercise, sleep, work, stress, relationships, etc. all have an impact).
So, I'm a big fan of the "assume it works because the studies say it should" approach that you describe. It's not clean or certain but it's eminently practical, and it has the fringe benefit of giving you a lot of opportunities to benefit from placebo :)