It’s kind of funny that Pareto principle applies to your point. Aka right model for the case explains 80% of things you care about with 20% percent of effort.

I kinda like graphs as abstract structure to describe things. It basically has collection of points/facts and edges/relationships. And relationships are based on theories. What is interesting is that part of graph is redundant. You can derive some points based on other points and theories about relationships. But there is more. Due to nature of arrow of time some of those are causes and some are consequences. Depending on what you are trying to do, explain the past or predict the future, you can look for right sets of facts tobderive missing points.

Using that metaphor generalists have more opportunity to move closer towards hypothetical, single true graph, or multiple.

But here is curious question, does one model exist? Probably but does efficient and economical version of it exists?

As far as I can say from math… General solution does not seem to exist. Have you heard of Abel–Ruffini theorem, Gödel’s incompleteness theorems, Tarski’s undefinability theorem? They all in different ways show that general, practical, precise solutions to many important problems do not exist. You can solve only done concrete cases, or will need to sacrifice things like precision or speed of finding an answer, sometimes to a degree of impracticality.

Reading on mental models a can’t not notice analogies. Mental models are efficient solutions for concrete cases of some problems. When applied in those cases they are incredibly useful. But when applied outside of those bounds they are faulty.

It’s kind of a tradeoff. Efficient solutions for a niches because general model does not exist or is too complex to have practical value.

Are you familiar with those theorems? Would be curious to hear your thought on niches and trade-offs and feasibility of one mental model to rule them all.