However, WHEN they are well composed and led properly, they can be valuable sources of data. They are at least a step up from simple raw statistical data, if done well. imho.
They can surface critical nuances that statistical data often fails to reveal. They can offer clues of fruitful channels of thought and/or action. They can suggest fruitful strategies with larger populations.
But, yeah, they can be crap, too.
Operative word is WHEN. I can see that.
Id be interested in knowing how the questions are developed. How the focus population is determined. What the ultimate goals of the questions are. How can bias be kept out of the whole thing? Etc. To be the fly on the wall listening to the brainstorming going on in developing such an encounter! :)
However, what you said about the critical nuances. Well... that would make it all worthwhile.
Just the thought of it, eh? (So hard not to pun.)