Partially Limited Characteristics (3rd followup)

[zornwil]

Why are derived characteristics treated differently than it was decided the final calc for Leap should be treated, i.e., as I recall, it was decided that the sum of leaps was not rounded at every step but rather the straight sum of every STR component and then divided? (Note - I'm not asking for or stating an opinion on right or wrong, just wondering what the rules reasoning is)

[Steve Long]

Because I think it makes more sense to do so given the possibility of differing modifiers, which I don't think affect Leaping or damage quite the same way.