How much self-knowledge is enough? argued against the idea of introspection as an optimal way to self-knowledge and I have expressed skepticism against the desirability of great striving to self-knowledge itself. My view is that is not necessary to know yourself precisely and that is also not doable. Instead, I think in terms of optimal self-knowledge. You don't have to know yourself any better than necessary. This view explains also my reservation against person focused interventions such as analyzing ones personality, identifying ones strengths, and person-focused compliments. I think this type of thinking reflects and plan-and-implement view on life and careers, whereas I find the validity of a test-and-learn view more credible. Instead of analyzing oneself and thinking about oneself a lot I think it may be wiser and provide more gratification to look into the world and find a way of interacting with it that seems to work both for yourself and for the world. You might call this 'doing what works'.
In Designing Positive Psychology, Michael Robinson and Maya Tamir have written the chapter A Task-Focused Mind is a Happy and Productive Mind: A Processing Perspective. They compare two modes of processing: task-focused and self-focused. Because of the limitation of our attention, these two modes of processing inhibit each other; so either you are self-focused or you are task-focused. The authors' proposition is simple: a task-focused mode of processing is generally more conducive to positive affect, mental health, and productivity while a self-focused processing mode leads to negative affect, psychopathology, and lesser task success. Sounds shocking? The authors not only link their idea to well-established psychological theories but also provide much empirical support.