I think that the condition to fear is over optimization and complexity doesn't make a system more or less stable. If redundancy is in place very, very complex systems can be robust.
So I guess over optimized systems with no redundancy is the danger.

There is complex, and then there is complex.

A mechanical watch is complex, but most of that complexity outside of the watch: i.e. the knowledge needed to produce a watch - it's not just counting teeth on gears, there's friction to deal with, shocks, temperature and gravity impact on the movements of the balance spring, and so on. A reasonable mechanical watch can run -20/+40 seconds a day.
Building watches requires highly specialized workers, high precision turning and milling tools, and a clientele able to pay for the work and the materials.
Watches may be relatively sturdy, the system of resources, work and education that is needed to produce such watches is much more fragile.
Today, it has become hard to find a decent mechanical watch that costs less than a month's salary.

A natural and functional ecosystem is also very complex, and composed of complex subsystems. The thing is this, in nature every subsystem has to coexist with local conditions. Also, ecosystems have high redundancy (highly diverse populations - insects, birds, mammals - can feed of the fruits of a tree, but it still manages to reproduce)
Our economic and social systems, however, are as 'efficient' as we can make them, and we are not satisfied with local inputs only. So we criss-crossed the world with trade routes, and pumped billions of barrels of oil into building our very affluent and increasingly indirectly connected lives. And all of this enterprise is based on more or less sparsely localized and finite resources. It is a cinch, that collapse is imminent for most of civilization.

Moreover, we have built our cultures on fiction. There is a lot of good stuff in there: the bible, the koran, Adam Smith, Karl Marx, all contain precious nuggets of knowledge and wisdom, but they are not science. They are incomplete and sometimes clumsy attempts at describing the human condition. Stories we invent about ourselves.
And then we started to really believe in the ultimate goodness of growth. In forty years of perusal of the press, I cannot remember a single issue of a newspaper that did not cheer or worry about growth, somewhere in its pages.
We just don't get it. If growth doesn't stop, the bottle explodes.
The flaw in the human system is, that it does not stop growing, even as outlying areas are being sucked of life. And the human system is imposing uniformity, reducing diversity.

So yeah, things can only get worse.

Personally, I have to admit that being conscious of this stuff doesn't change my life very much. I pay my debts, slowly, which fills me with misgivings, as I can see deflation looming. I go to work, and have a beer afterward. I tell scare stories to anybody who cares to listen. I sort my garbage, but my food comes from chain stores.
From where I live in town, it takes more than an hour to walk to the nearest field.
I use public transport and a bicycle, have no car. I have two adolescent kids who I try to warn about the coming dearth of everything. They spend a lot of time in front of screens, manipulating buttons. As do I.
Apart from trying to build a little stash of food, I have no idea of what I will do wtshtf. I would like to own a plot of land, but in a densely populated country, land is at a premium, and I have to get out of debt first.
I feel about as powerless as my neighbors, most of whom don't have a clue of what is hanging over their heads.
I'm looking to find a decent mechanical watch. If I muddle through, I would like to have a memento of our technical prowess to bequeathe to whomever is left.

I agree with Joule.The lacking definition is "complexity". For us, it is complex a system that we do not easily understand. Another definition is "chaotic", that fractal mathematics describes as a behavior that we do not easily understand or that is largely disturbed by unknown variables...
In a strict sense all systems are complex and chaotic, but we are adept of over-simplifying things to get a proper mathematical description. Take a basic example of dropping a weight from 1 meter height, simple physics, isn't it? We can reduce the formulas to Fg=m.g.h, f=m.a, and calculate the uniform acceleration movement to get the time of flight. Obviously we will get it wrong, and no measurement will result in the calculated value, because we did not take into account the air resistance,derived from the shape of the object, that depends of the orientations it takes in flight, and so on...
The worst example possible is drop a sheet of paper, we cannot even know the position on the ground it will end up in! It depends on to many factors, the way it is dropped, the winds, the dynamic deformations of the surface, etc... So, if we cannot get these simplest thing right, how can we expect to understand the behavior of huge systems like our own biosphere?

There is no simplicity, except the one we made up in an abstract and imaginary world, that discards variables that "appear" to have little influence. But some times we discover that the "little influence" is not so insignificant after all.

What we are missing is the humility to understand our ignorance...

Exactly right, accelerating growth on top of dispersion is what gives rise to the Logistic function leading to the characteristic Hubbert bell-shaped curve. If it wasn't for the accelerating term, the dispersion effects combined with proportional draw-down would flatten out the production curve over a much longer time period. We probably wouldn't even discern a distinct peak.

What would have happened if we had good analytical methods such as dispersive discovery and the shock model long ago? These do not involve difficult math and the use of computers is not necessary. I assert that careful watchers of the evolution of resource depletion could have fit the models to the empirical data early on and applied a kind of continual Bayesian update to monitor the situation. All the warning signs would have been clear much earlier. The heuristics of Hubbert were simply one man's intuition into what was to come. However the heuristics had no fundamental basis and as more data arrived, the adjustments necessary to update the heuristics had no justification and critics pointed out (arguably correctly) that early Hubbert followers were "shooting in the dark" about projections.

People question whether this is all water under the bridge given our current situation, yet I say another vital resource may come along that we can monitor more carefully.

Just a note [with full credit to TODer Paul_the_Engineer] that Zinc might be a future Liebig Minimum, too. Peak Everything sure seems to be hitting everything at once, that's for sure [but water shortages will be the absolute worse, IMO]:

http://www.theoildrum.com/node/5491/511539

Deleted Duplicate

As an aside, is anyone else frustrated that the only google results for the "Porridge Principle of Metered Decline" and "Asimov's Bio-Elemental List" point back to this site? I'm trying to pull together some real logic from this "contiuing thought process" and it's not easy.

I know that a google failure doesn't mean it can't exist somewhere, but the link you gave for the Asimov list in another post (http://www.backyardnature.net/phosphor.htm) doesn't mention a list, nor does it cite the source of the Asimov quote.

I don't want to start a flame war; I know Mr Shaw is a long-term member with high standing and a lot of respect here, but this isn't about Mr. Shaw, it's about the quality of this particular post.

Following the links for each, I can eventually trace the "Asimov's Bio-Elemental List" as perhaps a badly cited reference to an essay titled "Life's Bottleneck" from April 1959.

So not only is the primary source not properly cited, it's from a source by a scientific thinker (even fantasist) over fifty years ago. The only claim that stands up here is that phosphorous is needed to sustain life.

You haven't cited a source to justify your claim about a "Twenty-Fold increase in [S]ulfur", and you don't even say if you mean production, use or price. Later you say price decreased, does that imply more supply?

It may well be that the site is really all about a small closed community who all are supposed to know what an Earthmarines and what a Webb/Pomerene is, and why Directed Decline is different from directed decline. It may be that I'm supposed to spend a few hours researching all the "She comes down from Yellow Mountain..." previous posts in order to try and figure out Bob's point here, and if this is the case please tell me and I'll pull my head in.

On the other hand, if this post is supposed to communicate some sort of idea or concept to an intelligent educated reader, it fails to do so.