Some guidelines to keep in mind when collaboratively getting feedback on "soft skills" practices, that I think would also be effective for meditation practices.

In the coming years, it will be interesting to see if the teachers of various LGAT's  - from the big ones such as Tony Robbins, to smaller ones such as various Qi-Jong training, or "embrace the inner male" weekends, or whatever - start to embrace a collaborative and transparent approach in the feedback component of their various workshops. (This can even be utilized for meditation retreats.)

To the point where each particular workship or practice will have open surveys, fully publishable and searchable, on the web. 

Once we have put together (and professionalized) the various open source components (the survey tools, the LMS, the CMS, etc), the individual development paths, we can make the environment so attractive that the target audience of self-improvement/spiritual/community focused, will flock to the site.  Most workshop teachers/practitioners will have to embrace the model, as a cost of entry/advertisement.  This will have a self-reinforcing loop of improvement, as the best teachers will share openly results for practitioners, and improvements will come naturally out of that.

The basic premise behind open source:

"With enough eyes, all bugs are shallow" - Eric S. Raymond

Two premises behind Integral:

AQAL - all quadrants, all levels

"Each valid mode of knowing consists of an injunction, an apprehension, and a confirmation." - Ken Wilber

Actually, that last one isn't so much Ken Wilber's but a clear restatement of what has been understood in some schools of  philosophy for quite awhile.

These three are the fundamental positions necessary to begin a collaborative, but deep, investigation  into related realms of knowing, being, feeling, acting.

But the questions becomes - how to collaboratively "measure" or evaluate goodness or truth, in a less defined field than programming? 

In programming, you can both "see" well written code, in the actual programming, and you can also immediately see the effect when you run the code.  Instant feedback.

To collaboratively measure the result of practices not so easily seen, what is the feedback that is relevant?  What can be seen?

I would suggest we start with rich, well-thought out surveys.  If you look down to the Quality of Life survey, you will see that these type of surveys, given the quality of the questions in a survey, is the primary mode to enable quality feedback in a collaborative quest to know the truth, in human realms of practice. 

And, as I mentioned previously, the software exists to support the generation of good surveys.  We simply need to enable this correctly.