We report calculations for electron transfer rates at immiscible liquid interfaces with new models closely related to a model originally developed by R.A. Marcus. We compare our modified models both with the original Marcus model and with an adaptation of it by N.S. Lewis to electron transfer at the semiconductor-liquid interface. On the basis of these comparisons and a thorough examination of the foundations of both the Marcus and Lewis models, we find no basis for the assertion in the literature that the Lewis model can be used to calculate an upper limit for the electron transfer rate constant at the semiconductor-liquid interface. Features we discuss of the Lewis model and a related model by H. Gerischer show that these semiconductor-liquid interface ET models are evidently not physically plausible. We also conclude that while the Marcus model appears to be a physically plausible first order model for immiscible liquid systems, in its present form it does not seem that it can produce reliable quantitative estimates for real liquid-liquid systems. Our enhancements, both suggested and implemented here, of the Marcus model appear to offer the potential for constructing augmented and better defined models for immiscible liquid ET.