Cooperative driving automation (CDA) enables connected and automated vehicles to cooperate with surrounding vehicles and infrastructure for increased safety, mobility, and energy efficiency. CDA systems are categorized into four classes, depending on the cooperation level: status-sharing, intent-sharing, agreement-seeking, and prescriptive cooperation. In order to maximize the benefits of these systems, new communication frameworks and protocols need to be designed based on extensive studies on corresponding vehicle control performance. This work investigates the potential energy savings from using different CDA classes in car-following scenarios. The essential parameters of control and communication for reliable control performance and real-time implementation are identified, such as agreement-seeking frequency, prediction horizon length, and the number of CDA participants. In addition, important control design factors that need to be considered in CDA development are discussed, including the smooth transition between cooperative and individual driving plans and the proposals that maximize the probability of agreement from counterparts.