This shows a broad spectrum of proponents and critics of the approach: https://www.google.com/search?q=five+whys+site%3Anews.ycombi...
In reality, complex systems are driven by networks of causal effects, where some form reinforcing and balancing feedback loops. There's no start or end to a causal pathway, there's just where we choose to look and what we choose to ignore. There is no root cause, there is a myriad of interlocking factors that together dynamically drive the system in certain directions.
Five whys type analyses tend to end up blaming whatever is convenient or culturally acceptable to blame, and leaves many branches of the causal network underexplored.
Five whys is easy to explain and humans love the narrative point of view promoted by a chain of events, but it's a very inefficient way to explore the causal structure of a system.
"Five Whys," is a variant of this game where you simply roll the dice five times.
If you want to put on a show and dance to let your management and colleagues know you care about quality and are doing something about it, then it's great. If you like false assurances and feeling like you've contributed to a system you cannot fathom or understand it will definitely give you warm tickles. And it makes you sound smart when you put on such an impressive display of initiative and critical thinking.
But could there still be some value to this sort of exercise? At least it terminates at something. If it is possible to get rid of that convenient or culturally acceptable blame attractor, at least one problem has be exorcised from the organization, right?
That being said, the process at least forces the team to look at root causes beyond what is visible on the surface. It's a good start but it isn't perfect either.
What 5 whys helps to do is to structure your existing knowledge. In my experience it is great when you want to analyze your own decisions, or structure team knowledge about particular issue down to a necessary level of detail.
For actual root cause analysis there are better tools, e.g. Ishikawa[1]
Edit: oh, because the link actually has hacker news in the query filter... duh...