What about metal carbonyls? For instance, hexacarbonylvanadium, drawn with charge separation (c), looks really ugly. On the other hand, the software (e.g. ChemSketch) objects to the representation (d) (which, apparently, is preferred; cf. tetracarbonylnickel on p. 408 of IUPAC Recommendations) because it ‘wants’ the positive charge on triple-bonded oxygen.