I think Mithoron has the right idea. Say before chelation your cation is in water, surrounded by its typical hydration sphere of solvent molecules coordinating to it. Then the chelating agent comes along, displacer the water. I'm guessing the driving force for the substitution is enthalpic in origin, the cation-chelator coordinate bonds are more energetically stable/preferable.
Intuitively, which system can be arranged more ways: 1) Bulky chelator molecules strewn about a sea of solvent molecules, and metal cations coordinated by some number of solvent molecules, or 2) that number of solvent molecules freely dispersed in the solution, and the chelators coordinating the metals...?
Since the free chelators themselves require an ordered solvent structure around them, my intuition suggests that performing the chelation and in the bulk solution swapping chelators for solvent molecules would increase entropy, because these newly freed solvent molecules can occupy nearly any position and orientation (whereas previously the free chelators required solvation spheres).
tl;dr Perhaps the entropy increase of dissembling the ordered solvation spheres around the free chelator molecules compensates for your entropy decrease of forming the ordered chelator-cation complex, leading to a net increase in entropy.
I think your situation may be interesting to compare to the rationalization of the hydrophobic effect.