manganese is a B-group element, and you can consider permanganate as to be a B-group element complex.
many of those B-group element complexes are coloured in VIS , due to , for example, eg-t2g electron transition effects: the energy required ( i.e. 10 Dq) often is in the range of VIS light-energy
from another point of view you could call it the HOMO-LUMO energy difference that is suitable for those effects with complexes of B-group elements
... and exactly this (mostly) is NOT true for maingroup "complexes" (if you regarded carbonate to be the oxygene-complex of carbon): the 10 Dq there (there is one, too, yes) , i.e. the HOMO-LUMO difference is much too large to be reached by VIS
... and hence carbonate appears to be colourless
regards
Ingo