The only "logic" is that we don't have large numbers of aircraft running off the sides of runways when an engine fails; the conditions for Vmcg are arbitrary, and they way we apply that restriction to the performance data is arbitrary. But the end result is an acceptable "level of safety".
It's the same as trying to find an academic reason for using 1.2Vs or 1.13Vsr; there isn't one. It just happens to work out ok most of the time. If we had 12 fingers we might
still be using 1.2Vs, just because we'd have picked the digit "2" for the first "decimal" (twelvical??) place; but it's be the equivalent of 1.167Vs in decimal notation. If it works....don't amend it.
