The adducts between luminescent lanthanide tris--diketonates and diimine or triimine ligands have been explored exhaustively for their exceptional photophysical properties. Their formation, stability and structures in solution together with the design of extended metallopolymers exploiting these building blocks remain however elusive. The systematic peripheral substitution of tridentate 2,6-bis(benzimidazol-2-yl)pyridine binding units (Lk = L1-L5), taken as building blocks for linear oligomers and polymers, allows a fine tuning of their affinity toward neutral [Ln(hfa)3] lanthanide containers (hfa = hexafluoroacateylacetone) in the [LkLn(hfa)3] adducts. Two trends emerge with (i) an unusual pronounced thermodynamic selectivity for mid-range lanthanides (Ln = Eu) and (ii) an intriguing influence of remote peripheral substitutions of the benzimidazole rings on the affinity of the tridentate unit for [Ln(hfa)3]. These trends are amplified upon connecting several tridentate binding units via their benzimidazole rings to give linear segmental dimers (L6) and trimers (L7), which are considered as models for programming linear Wolf-Type II metallopollymers. The modulation of the affinity between the terminal and central binding units in the linear multi-tridentate ligands deciphers the global decrease of metal-ligand binding strengths with the increasing length of the receptors (monomer  dimer  trimer  polymer). Application of the site binding model shed light on the origin of the variation of the thermodynamic affinities: a prerequisite for the programmed 2 loading of polymer backbone with luminescent lanthanide -diketonates. The analysis of crystal structures for these adducts reveals delicate correlations between chemical bond lengths measured in the solid state (or bond valence parameters) and metal-ligand affinities operating in solution.