To understand something of the conundrum faced by teachers new to pronunciation work today, you probably need to begin with Kumaravadevilu's (2003) initial characterization of the "Post-method" era, or, even better, Kumaravadevilu (2007). That, along with a great new book just out, Pronunciation Myths, edited by Grant and Brinton, provides a good perspective on the question.
In essence, the first (a) builds a compelling case for the idea that no one method can possibly work in all contexts and (b) provides a set of 10 strategies for, in essence, building your own personal classroom teaching method--while at the same time warning that an inflexible "method" is to be avoided at all costs. In part that is because the only way you can test a method is to try it yourself, in your classroom. And also you must--at least temporarily--believe the "testimonies" of those who use it, while you test it.
The second, "Myths," while thoroughly dispensing with several common misconceptions (and a few "straw men") about pronunciation teaching, provides a very useful review of the range of research-tested techniques that have, in fact, been shown to be effective. (I count 40 or 50 discrete techniques or variants in all.)
Add to those two the current theoretical perspective that pronunciation must to the extent possible be integrated into general instruction and you have the post-method conundrum: How do I personally assemble the techniques that can be integrated and will work in my (unique) classroom? (Murphy's chapter in "Myths" addresses that issue quite well in fact.)
What we refer to as "Haptic-integrated pronunciation teaching" is, in fact, a method based on the use of a wide range of well-established and proven techniques. (All reported in "Myths!") What makes it different is just that (a) the techniques are generally anchored (or reinforced) using movement and touch--based on multiple "haptic" studies in other disciplines, and (b) the method focuses principally on modelling, feedback and correction--and to some extent integration into spontaneous speech. Those are dimensions of pronunciation teaching that have been studied extensively in "the lab" but not in the classroom, especially in terms of long term improvement.
And to finish up the post-modern stew: EHIEP (Essential Haptic-integrated English Pronunciation) work is essentially experiential, both learning with it and about it. The best evidence that we have now that the method, itself, "works" are about a decade of reports from students in the classroom and related research from a dozen other disciplines. The same is the case with all methods, of course.
For a number reasons, it is exceedingly difficult to test a method, among them the fact that no matter what the results, in today's "hyper-localized" theoretical view of methodology, the nature of the learner population may radically limit generalizability. Few if any classroom studies or action research in pronunciation teaching provide much in the way of detail as to how the techniques or treatments were actually conducted or relate to the other instruction or ongoing experiences that students/subjects were involved in at the time of the study.
Theoretically, one should, of course, be able to generalize from the local. (That is, after all the raison d’etre for the dominance of qualitative research in the field currently.) In practice, research is today still so thoroughly “critical-agenda-driven” that general applicability of methodology cannot be of interest. In time, in depth studies of one teacher's method will again be in fashion and doctoral dissertations. (I'm working on a book proposal that will do something like that.)
For the time being . . . Just do it!