It may work on paper--though I'm not sure why the theory would suppose it's a good idea to retire a really good leader so someone else can get some experience; it may even work sometimes in practice, but most of the time in small teams leadership is not settled and effective until "dominance" is settled.
Say what you will about one for all, and all for one. There's always a dominate figure. Dominance is not synonymous with being a bully. Dominance is a $10 word for pecking order and the ability to project (as in projection) "presence" (as in command presence).
Sharing leadership (in the sense of passing the baton) is often difficult since the pecking order is not easily rotated. Some folks are natural followers, others are managers (task oriented) but not leaders (message and concept, direction, protection, order)
Hersey and Blanchard told us all about this years ago with situational leadership. Most people who know about this (they used to teach it routinely in management school) remember the four S's for the four situations of leaders. But the more important point about the SL model is that is posits a place for followers.
Followers have styles also; some can be very assertive, but nonetheless managerial rather than leader. The idea of SL is that the leader--being the adult--adjusts his/her style to be the compliment of the follower: assertive follower--> delegating leader, etc.
It's true that on some larger scale than a small team, social and institutional emergence is at work. In the event, we can expect self-organization. In the United States, we need only look to the assembly that that put together our Constitution to see how this works. Of course, that self-organization was then ratified by the affected population.
But I'm talking about something on a smaller scale: a dozen people forming themselves first into a working group and then into a committed team. By the time you get to the team thing, dominance will have worked it's will: a leader will have emerged. Managers in the team will have suggested the right organization--that's the self-organizing thing, and good managers can make it work (managers are better at this than leaders, who may have their head in the iCloud or some such). Task leaders and the team leader will then figure out the right styles... and off we go!"
And, in what direction are we going? Should we follow the customer, lead the customer, or ignore the customer in the short run? That last thing, ignore, is not all bad, because as Henry Ford said (paraphrase): if I had asked what they wanted, they would have said "faster horses"
The answer, of course, is "it depends..." (And, it always does).
