The 80:20 rule is a lot like the Golden Ratio, in that it is easy to confuse coincidence with causality. Maybe the fact that bosons that are freezing their metaphoric asses off tend to collapse roughly 80% of the time by volume is indicative of some unfathomable force at work in the universe, but the fact that the ratio of genus to species classification follows this rule is probably more a function of enforced simplicity than metaphysical mystery. By the same token, things that follow a normative, expected aesthetic proportion that is encountered throughout nature as a fundamental standard will undoubtedly be somewhat pleasing by default...but that doesn't mean your lotto ticket will be more likely to pay out if you use a Fibonacci sequence to pick your numbers.
But these tendencies, even if not universally applicable, can be used to model a wide variety of things. Failure distributions, root cause analysis, and even "people involved who complain" certainly seem to adhere to these rules in my own line of work, and the concept of a Pareto distribution as explained by Chris Anderson's "Long Tail" model for online businesses is demonstrably correct...so correct in fact that people are starting to utilize it deliberately now to model a strategic effort that is designed to cater to the disparate masses rather than the focused few that drive 80% of the business.
In politics, it is easy to apply the same logic; what is "big tent" politics or "coalition building" if not offering a home to the many specific special interests that occupy that long tail? Parties reach out to the multitude of weirdos on the fringe because it is easier to polarize them on very specific values and draw a distinction based on that limited argument than it would be to convince them that they all had at least something in common worth working for.
Ultimately, the problem is that in latching on to the model most adoptees have forsaken the intent of Pareto's original work in economics, which was to identify optimality.
