From your question, I guess you're coming from the theory point of view. It's being said more and more. Theory attempts to explain what's happening. NOT theory came first, so things ought to fit into its parameters.
If the latter was the case, we'd still be playing early madrigals; Bach, Beethoven, Schonberg, Debussy, et al, would never have emerged. Surely, the main reason we feel that something works is that it sounds good. It isn't always possible to pigeon-hole particular aspects, even though we seem to be programmed to need it done.
The main 'reasoning' here is that music needs (to us) to resolve, and moving each note in a chord as little as possible is a good way for this to happen. Almost like - you know the sounds you're expecting, but the chord before is so very close, but not quite there. A bit of a tease - dissonance, maybe.
C E G Bb Db. To resolve to F A C E involves C and E staying put, but the E can move one semitone to F. Bb coming down one semitone to A. Db dropping one semitone to C. G moves a whole tone - unless a jazzer would play the F as Fmaj9 ! Very little movement all told, but the blend of penultimate chord notes sounding good.