The properties of polymer networks depend on their connectivity, specifically on the presence of connectivity defects like loops and dangling ends. While chemically crosslinked networks have a frozen distribution of defects on multiple length scales, equivalent transient ones are capable of self-repairing. To study this potential, we employ spatial hindrance as an effective self-sorting mechanism to develop metallo-supramolecular polymer networks based on heteroleptic complexes. We functionalize tetra-arm poly(ethylene glycol) (tetraPEG) chains with the mesitylene-substituted phenanthroline ligand, which is incapable of homoleptic complexation, yet it can form heteroleptic complexes with tetraPEG precursors functionalized with unsubstituted phenanthroline or terpyridine ligands. This way, the formation of primary loops can be avoided upon selective formation of heteroleptic complexes. Nevertheless, the coordination geometry preference of the utilized metal ion can tune the extent of self-sorting and consequently the network structure and defect composition. The latter is pursued via MQ-NMR, and periodic density functional theory simulations are used to visualize the structure and explain its relationship with the dynamics as measured by rheology. Our results demonstrate that Cu(I) can inherently adopt the geometry requirements of both desired heteroleptic complexes by benefiting from the πstabilization effect, thereby creating a network with the least primary loops and dangling ends.