You said that n can be flexibly defined in different forms for each individual node. Could you explain what specific kinds of additional constraints or variables this flexible definition requires for predicting stability in large-scale systems?

For each node, the challenge will probably be how sophisticated the model is in capturing n, that is, how many people or how much scope of work each manager can actually manage well. There are many things to consider in this complex simulation: whether that capacity is innate or grows with experience, and if it grows, what kind of learning curve it follows, and so on.