Regarding the method itself, I have some doubts: too much depends on how exactly you answer the "why" question. Minor differences in wording lead to different directions of the next "why"-question. And thus you might arrive to a totally different root cause.

Agree. I have seen (and even participated) in this process. More often than not, this turns into "ok let's find some questions and answers for this paperwork so we can get back to real work". 5W is certainly a good tool to start training your team to do deeper investigations, but it is by no means a full solution for backtracking issues.

The inherent assumption of the method ("there is only one (relevant) root cause") is IMHO quite questionable. In my experience, there are often multiple reasons that "come together". Of course you can tackle this by applying 5why to various bones of your Ishikawa-diagram...

But the fishbone also gives the impression that each of the different root causes don't interact with each other until the end, if i understand that one correctly.

Reality is often times much more messy, with some issues slowly making each others worse over time while others are new and only come into play at the very very end, or have lingered harmlessly for years or decades until the circumstances changed. Circular reasoning and things interacting (or not interacting) in unexpected ways under unexpected circumstances can also play a role.

And a single long standing root cause can split into multiple pathways of issues that only turn into the final failure when they meet again.