Cloud simulation toolkits such as CloudSim and GridSim are widely used to evaluate resource provisioning and scheduling algorithms in virtualised computing environments. However, their standard models are constrained by architectural assumptions centred on finite workloads dispatched to virtual machines or grid resources. A systematic assessment is conducted to determine whether these frameworks can faithfully represent the BlackAnt platform — a production cloud execution environment built on Docker Swarm-orchestrated microservices, a replicated control plane, task-arrival-driven auto-scaling, and OAuth2-secured HTTP REST communication. A structured gap analysis across eleven architectural dimensions shows that the frameworks do not provide the first-class semantics required by BlackAnt, including long-running replicated services, service discovery, asynchronous application callbacks, and persistent distributed control-plane state. The minimal abstractions and an implementation and validation roadmap for a successor simulation framework are characterised.