Two-phase flows in polymer electrolyte membrane fuel cells (PEMFCs) are complex dynamic processes involving phase transitions and phase production due to concurrent processes. There have been attempts to simulate these phenomena using the multiphase mixture () model. The conjecture is that is mathematically equivalent to classical two-fluid models without invoking any additional approximations. We show that the model has a narrow applicability, limited to flows without phase transitions or phase production due to other processes. For more complex situations, including those encountered in PEMFCs, the model ceases to correctly reflect the conservation principles and may lead to predictions of unrealistic velocity and scalar fields.