The length of a second is entirely arbitrary. It is an accident of history that the second is what it is. When it came time to set a more empirical basis for the length of the second as our best clocks had already come to measure it, the best we could do is pick an atom that oscillates at a rigidly consistent rate, and determine how many oscillations happen in a second as we have come to use it. That number happens to be 9,192,631,770 oscillations of a Cesium atom between two of its energy states, which I think we can all agree is a rather arbitrary and meaningless number in itself. There was nothing more elegant available, and certainly nothing that could be based upon the astro-chronology of earth itself, for the length of the year and of the day slightly changes over time.

What this means, is that the length of the second as we have come to know it could just as well be longer or shorter than it is, and if so, 432 of 440 A would be a completely different note. Or to put it another way, if our second were, say, just a wee bit longer, 440A as we know it could actually BE 432A as measured by our hypothetically slower second.