How realistic is the connectivity in a testbed? The question determines whether a protocol that performs well in a testbed also works in a real-world environment. We answer this question by gathering 11 datasets, a total of 2,873,156 connectivity measurements, on both testbeds and real-world deployments. We determine that real-world deployments feature some quantifiable level of external interference, multi-path fading and dynamics in the environment. We further show that some testbeds do not have all three components. We develop a 5-point checklist to assess the realism of a testbed deployments and introduce the visual "waterfall plot" tool to instantaneously evaluate the connectivity characteristics of a deployment.