This paper develops and validates a parametric model to explain the performance benefits of vector tracking loops over their scalar counterpart. It also presents a systematic methodology for comparing scalar and vector architectures with similar design parameters (e.g. discriminator types, loop update rates and loop noise bandwidths), and thereby identify conditions under which both architectures can perform comparably. A theoretical analysis of the parametric model backed up by simulation studies offers a different look at the workings of the vector architecture. It shows that internal aiding among loops enables better performance of the vector tracking architecture. So, this suggests that well designed scalar tracking loops with external Doppler aiding can offer similar performance improvements, while avoiding the complexity of the vector architecture. Additionally, internal aiding is also found to be responsible for easy propagation of errors among loops and, therefore, affects the fault tolerance capability of the vector tracking architectecture.