Berin Loritsch has wise words on action vs. planning, notably:

The less experienced you are, the more you need to experience (I think he means experiment) so that you can learn strategies to deal with the mistakes you have made.

To which I'd add: the less experienced you are about your particular problem, the more you need to experiment.

Projets where you're tackling new stuff or where (as is often the case) requirements are unclear benefit a lot from agile processes. It's not only the people's seniority that comes into play, but also the "novelty" aspects of the job.