This paper provides a detailed discussion about currently proposed Elliptic Curve Cryptography (ECC) models focusing on performance parameters such as Security, Complexity, Scalability, and cost of deployment. It was observed that Machine Learning optimizations including bio inspired computing, deep learning, and transformation models outperform other techniques. This discussion is extended via an empirical estimation of these models related to the performance metrics under different application scenarios. This paper also proposes the calculation of an ECC Performance Metric (EPM), which combines the evaluated parameter sets to identify ECC Models that can perform better under multiple operating scenarios.