Well, the short answer is that these rules of thumb are also described in other textbooks. But (if I remember correctly) these rules are just made up by induction of the principles used by old composers who did use rules but these were the rules of counterpoint and not of voiceleading. Voiceleading is in principle not much easier than counterpoint, but in practice I'd say it is.
These rules of thumbs are, of course, no actual laws. You are free to stretch their boundaries, or leave them altogether. Modern composers don't use these principles of voiceleading too often, but even composers like beethoven and mozart and bach etc did not always totally adhere to them - especially in instrumental (keyboard)music. Basically, these rules are a way of making sure that (in classical composing) the voices and harmonies shift as smoothly and as pleasant as possible, and making sure that every voice is very much independent and moves stepwise and logic (that is the why for those rules regarding crossing of voices, paralles etc). The rules are based on vocal music, and because vocal music has it's limitations (the pitches have to be created by the voice, not just by putting a finger down on a piano or violin) people wanted to write for it in such a way that it would not give any trouble when singing.
Those 2 rules you stated are ways of making sure that all the voices move as little as possible between chords, because large leaps are harder to sing than small leaps. And they also make sure that no voice is deprived of its 'independence' to the others: if two voices move in unison or octave, they are practically doing the same, therefore making the distinction between the two hard/impossible. I think this fascination for independence came with the polyphonic practices and stayed even when composers were beginning to write more monophonic.
The 5th is doubled if the root cannot or isnot to be doubled. Just make sure that the leading tone (7th tone in a scale) is never doubled: it is too dissonant for that (according to the rules ofcourse).
'More satisfying' would mean that the total amount of stepsize is smaller, that some form (like a cadence) is used, or that they just voted before sending the book into print on what voiceleading between chords would be most agreeable - and that one won.
EDIT: Ah yes, to the how and why for specific exceptions you should listen to someone like siwi