The Competency Zone
In this zone, the difficulty of a task and the competency in executing that task are largely in sync. In learning terms, sufficient knowledge has been acquired to allow a clear demonstration of some level of ability—this qualifies the Zeroth Order Ignorance (0OI) criterion of provability. Without being able to clearly show competence and knowledge one might well be self-deceived and subject to the Dunning-Kruger syndrome. If a computer system has not passed its required tests it has not demonstrated that it was built "correctly" and almost certainly will not work as required.
In the "higher levels" of the Competency Zone there is likely to be holes in the knowledge and ability. That is, the person is competent to a certain degree, but still has room to acquire some knowledge and skill.
While the graph shows a clear delineation between these zones, obviously in practice it is a gradient. The lack of knowledge might be of specific components of the needed knowledge or skill: a tennis player might have a wicked serve, but a much weaker backhand—something a savvy opponent would take advantage of. A computer system might process routine transactions perfectly, but throw errors on some odd combination of conditions. In this case a savvy developer might pre-filter or shunt aside transactions that might trigger these errors.
The Comfort Zone
This name is most appropriate when referring to humans. Having clear and apparently effortless competence is a very comforting feeling for most of us. It is the low-energy state where, in the extreme, the task can require little to no conscious attention and mindfulness. Since it is positioned at the low end of the Competence Zone, it seems there is not much that needs to be learned, which makes for even less effort
Computer systems neither value comfort nor experience the boredom that the Comfort Zone implies. So perhaps the "comfort" is experienced by the maintenance programmers who do not have to work weekends to fix (read: add knowledge to) the system because it broke (read: did not have enough proven knowledge in it). While working as a maintenance programmer on high quality stable systems that never go wrong might be boring, I have personally never heard such a complaint from any such programmer.
In the Comfort Zone, there is little pressure to learn anything new; indeed, there is often pressure to not expend the effort to learn anything new. Any learning that does take place will tend to focus more on further reducing effort rather than actually expanding the knowledge base—additional comfort trumps additional learning. Certainly, there is no need for the wholesale replacement of knowledge structures that require a "paradigm shift"—the current level of knowledge is just fine.
Why Move Out of the Comfort Zone?
The Comfort Zone is, well, comfortable, so why change?
Any movement out of this zone is usually triggered by external events. In the Dunning-Kruger model, the increasing and unavoidable proof that someone is incompetent (despite having thought otherwise) occurs when the external world (representing reality) imposes its will. It could be viewed as an involuntary vertical adjustment along the Difficulty axis—the task turns out to be much more difficult (ie., requires more knowledge) than was initially and incorrectly thought. Alternatively, the movement could be viewed as a regression backward along the horizontal Competency axis—faced with reality, the skill (read: knowledge) present is shown to be insufficient. I think in most cases both happen. Faced with a true learning requirement, we find that (a) it is harder and requires more knowledge than we thought at first blush and (b) we don't know as much as we thought we did.
Now, to actually become competent, we are forced to learn.
For some self-actualized individuals, slipping into the lower reaches of the Comfort Zone is boring enough by itself to galvanize them into action. After a while in a mundane, routine job which the person can do with eyes closed, there is a need for more challenge. Staying in the same job doing the same thing over and over will not usually provide that challenge. After all the efficiency changes that can be made have been made, there is nothing more to learn. While some will be content with this, others will not. If the difficulty cannot be suitably increased, a change of environment and task is required. This change might be a professional one or the desire for greater challenge and more opportunity to learn may be manifest outside of the work situation: a machine operator learns to weld, an editorial proofreader learns a foreign language, a programmer learns to ski...
The desire to learn is inherent in all humans—unless the lure of the easy life in the Comfort Zone proves to be too strong. However, as Professor Csikszentmihalyi noted [1], the easy life also tends to be a rather unsatisfying life.
FOOTNOTES
[1] Csikszentmihalyi, M. Flow: The Psychology of Optimal Experience. Harper and Row, 1990