Pitfalls in the modeling of labor market flows: a reappraisal