Hi,
I would say that numerically, Telemac-2D's solvers are very robust and can go down to very fine horizontal resolutions, but the question should not be reduced to that only.
You must above all ask yourself what are the main processes that you want to solve for the purpose of your modeling project and how Telemac takes them into account: are they fully/partially solved, parameterized, neglected?
For example, turbulent processes require very fine scales (in time and space), which Telemac-2D does not completely solve (even with small grid elements and a solver convergence) and for which other models (3D, CFD ...) would be more suitable.
I hope it helps.
Regards,
Laurent