Traditionally there have been two distinct classes of mathematical or conceptual models to describe the behavior of entangled polymers: temporary network models and tube models. We have taken an approach that marries these two different views, using the idea of a slip-link. Key to this novel development is the use of monomer density between entanglements as a stochastic variable. The resulting model contains only a single phenomenological parameter (a time constant), which is determined by linear viscoelastic experiments (or possibly by molecular dynamics). All nonlinear flow dynamics are then described without adjusting parameters. This approach offers a way to model systems that have been modeled by separate equations, to date. The resulting theory has given excellent predictions of all known shear data, and steady elongational data. However, it has not been able to describe transient elongational data. We will examine and eliminate the use of simplifying assumptions in previous implementations of the theory as a possible cause of the remaining discrepancy with data. We will also show how the current theory can incorporate semi-flexibity, branching and even cross- linking without a significant change to the underlying mathematical structure, making it a useful bridge between atomistic and macroscopic descriptions for soft biological materials.