Abstract: Despite (or because of!) the diversity of SGS modelling techniques, the implicit LES approach is expanding within the community of high-fidelity turbulence simulation. The results obtained in this framework lead to legitimate questions about the actual usefulness of explicit SGS modelling. This question can even be extended, to some degree, to the use of a wall turbulence model. In this seminar, these two questions are addressed with the objective of identifying limits of the implicit SGS modelling approach. The discussion will be conducted on the basis of results obtained for two fundamental turbulent flow configurations: the 3D Taylor-Green vortex and the pipe flow. The analysis of a priori and a posteriori results will be an opportunity to question the LES formalism and our expectations regarding the results obtained in this framework.