We have studied emulsions made with two- and threephase oil-water-surfactant systems in which one of the phases is a microemulsion, the other phases being water or/and oil excess phases. Such systems have been extensively studied in the 1970-1980s for applications in enhanced oil recovery. It was found at that time that the emulsions became very unstable in the three-phase systems, but so far few explanations have been proposed. In the most complete one, Kabalnov and colleagues related the emulsion stability to the probability of hole nucleation in the liquid film separating two nearby emulsion drops and associated this probability to the curvature elastic energy of the surfactant layer covering drop surfaces. We propose a different explanation, linked to another type of interfacial elastic energy, associated with compression of the surfactant layers. As found long ago, the three-phase systems are found near optimum formulation (hydrophile lipophile difference, HLD = 0), where the interfacial tension exhibits a deep minimum.the determination of interfacial elastic properties in low interfacial tension systems is not straightforward. In our present work, we used a spinning drop tensiometer with an oscillating rotation velocity. We show that the interfacial compression elastic modulus and viscosity also exhibit a minimum at optimum formulation. We propose that this minimum is related to the acceleration of the surfactant exchanges between the interface, oil and water, near the optimum formulation. Furthermore, we find that the surfactant partitions close to equally between oil and water at the optimum, as in earlier studies. The interfacial tension gradients that slow the thinning of liquid films between drops are reduced by surfactant exchanges between drops and the interface, which are fast whatever the type of drop, oil or water; film thinning is therefore very rapid, and emulsions are almost as unstable as in the absence of surfactant.