Although the eFFBD formalism dates back to the 1990s (or even, in a simplified form, the 1950s), it is still not as much used by the Systems Engineering community as it could. As a tool for describing the functional decomposition of a system, it is indeed superseded by other languages such as the SysML activity diagrams or BPML. In this position paper, we would like to put the emphasis on the great potential of eFFBD as a credible alternative formalism for modeling, communicating and reasoning about complex systems. After shortly recalling the history and the key concepts of this formalism, we propose a brief assessment of its capabilities and compare them with other formalisms, such as Petri nets and SysML. We then identify several leads to improve the language so it better fits needs for the processes of system definition and system design. We then lay down the basis of xFFBD, a proposal for an evolution of eFFBD.