Should the Dicke model of light-matter interaction include a diamagnetic term? This question has generated intense debate in the literature, and is particularly relevant in the modern contexts of cavity and circuit quantum electrodynamics. We design an appropriate probing strategy to address the issue experimentally. Applying the tools of quantum estimation theory to a general Dicke model, we quantify how much information about the diamagnetic term (or lack thereof) is contained in the ground state of the coupled system. We demonstrate that feasible measurements, such as homodyne detection or photon counting, give access to a significant fraction of such information. These measurements could be performed by suddenly switching off the light-matter coupling, and collecting the radiation that naturally leaks out of the system. We further show that, should the model admit a critical point, both measurements would become asymptotically optimal in its vicinity. We finally discuss binary discrimination strategies between the two most debated hypotheses involving the diamagnetic term.