The need for slender structures in various engineering systems results in exposure to wind loads that can result in flow separation and re-attachment. This event is called a dynamic stall, and the resulting aeroelastic instabilities are called stall-induced aeroelastic flutter. However, a lacuna in the hitherto literature needs to completely describe the bifurcations in a stall flutter system and the underlying physical mechanism. This end of concern directly impacts the practicability of early warning measures to foretell aeroelastic instabilities. This paper employs complex recurrence networks to derive detailed descriptions of the bifurcating dynamics of this system. Precursor measures aimed at predicting the onset of flutter have been developed from statistical measures derived directly from the time histories. The system is numerically simulated using the Leishman-Beddoes (LB) semi-empirical dynamic stall model, which exhibits the required non-smoothness on a pitch-plunge airfoil. Through the same, we develop a suite of measures to foretell an impending transition in a stall flutter problem.