A street map leaves out the colour of every front door. That omission is useful when you need directions and troublesome if your task is to document painted entrances. The value of a model depends partly on the question it is meant to answer.
A model simplifies. The productive criticism is therefore not merely that it leaves something out, but that a particular omission matters for the use being proposed.
State the job before judging the detail
Write one sentence describing the decision or explanation the model supports. Then identify its main assumptions and the features it keeps. A diagram made to explain a mechanism may be unsuitable for predicting an exact quantity.
Consider a fictional queue model that treats every service interaction as equally long. It might help introduce the effect of adding a second desk, while performing poorly when a few complex cases dominate waiting times.
Test the boundary with a counterexample
Ask for a situation in which an omitted feature becomes important. In the queue example, compare an ordinary day with one containing several unusually long cases. The exercise reveals where the simplification may stop being helpful.
This is not a demand to add every possible detail. Extra complexity can make a model harder to understand and maintain without improving the intended decision. The question is whether the added feature earns its place.
Keep a short limits note beside the model: intended use, important assumptions and situations requiring another approach. When the model is shared, that note should travel with the attractive diagram or numerical result.
A useful model can be both incomplete and informative. Clear boundaries let people use it with judgment, revise it when the task changes and recognise when a confident-looking answer is being asked to do work the model was never designed to perform.