Reduced crewing concepts require a higher level of control and integration of platform systems. A clear reliability assessment of these systems in early design stages reduces the need for alternations in later design stages but remains challenging to perform. This paper addresses the design of reliable and integrated onboard systems such as cooling water, power distribution, and control systems. Current approaches to making platform systems more reliable, such as redundancy, modularity (independent subsystems) and reconfigurability, are analysed from a network theory perspective. Current graph measures do not align with experience-based requirements for improving system robustness. Our method combines the principles of network theory and experience- and rule-based system requirements to provide a comprehensive framework for a reliability comparison of integrated multilayer platform systems (distributing more than one type of flow). The robustness requirements are translated into network metrics to facilitate a quantitative trade-off typical to the early stages of the design process. The case study offers a preliminary view of the system topology of a notional naval vessel, consisting of power distribution, cooling water distribution and control systems. The network metrics facilitate an assessment of the system's reliability compared to alternative system topologies with differentiating numbers of nodes, edges and density. This study finds varying dependencies of the robustness metrics on the network properties, shining new light on whether and how one should compare distribution system robustness.