Product Line Engineering (PLE) is a paradigm for reuse-based complex systems development that is well installed in the industry. Among the proven benefits are reduced time to market, better asset reuse, and improved software quality [1]. To be successful, PLE must efficiently manage the variability -- the capacity of product line's artifacts to vary -- present in the products that form a Product Line (PL). Several modeling approaches have been proposed to represent the correct combinations of artifacts of a PL, their properties and relationships in a single Product Line Model (PLM) [2]. In this context, being able to reason about the PLM is an important success factor in the PLE strategy. Reasoning on PLMs is achieved by querying the models in order to verify, analyze or configure them [3]. For instance, PLMs can be verified to guarantee that they do not have undesirable properties affecting the correctness of the products they help develop. Several approaches are available in the literature to support automatic reasoning on PLMs [5]. Several approaches consist in transforming the PLMs into a constraint program that can be executed by a solver. For example, Satisfiability (SAT) solvers are used to analyze PLMs specified as Boolean constraints [6]. Others use SAT or constraint over finite domains solvers to find the number of solutions that can be configured on a PLM. Interestingly, it seems that for this task Binary Decision Diagram (BDD) solvers are more efficient [6]. Thus, authors seem to undermine the efficiency of certain reasoning operations to prioritize others. One reason might be that the transformation is guided by the solver to be used and not by nature of the PLMs or the efficiency/limitations of using one solver or another one.