How dramatic are the counterexamples though?

A browse through Morning Star's website reveals job listings for defined roles. They're looking for (i) people with "a strong business background" and "an aspiration for senior leadership" to manage their finances, (ii) people who "likely have a meaningful formal education in related disciplines" to "coordinate", "train" and "identify proper operating procedures", and (iii) "hands on professionals" who need the "ability to read, write, and comprehend instructions" in addition to specific practical skills.

Sure, they might downplay rank and authority, but it doesn't sound like an organization without structure or de facto managerial roles.

fair point. I think corporations live on a spectrum of authoritarianism, and these are "far out in the distribution" but they aren't perfect, and I don't think in reality a perfectly flat corporation can exist because all employees have a natural range of skillsets. But by "dramatic" I meant dramatic relative to the parent post's claim about organizational size relative to success of the model.

The Semco story is fascinating. I wish I knew some folks that worked there. The wiki (http://en.wikipedia.org/wiki/Ricardo_Semler) doesn't tell much but from what I remember of the book, they cut organizations free when they reach a certain size.

I used to work for Viant, a digital consultancy where they kept every office to about 100 (or was it 150?) employees. If they had to hire more employees, they'd spin off a new office. This was based on Dunbar's number (referenced elsewhere in this thread).

How did this work in terms of flexibility in the downturn? Did it help that the company was more nimble, or because of a lack of central control?

I'm curious because models that work in good times don't always work in tough times.

Sorry for the long delay in reply. Unfortunately, this "experiment" didn't reach a conclusion. The company ultimately folded.