Hey Ed,
Wow, THANK YOU for such kind words! So glad you liked it, and I'm especially glad you found the info to be useful. Hopefully other people will too.
As for your question, I'd say that it mostly comes back to what I was saying above: it's not about the chord/alteration (and, therefore, the upper-structure triad), out of context; it's about that chord's location within the key or key center, and whether such alterations would result in notes that are more or less diatonic to the original key or key center. It happens that
most of the usual polychord voicings in common use involve a triad of some kind located a third of some kind above or below the root. So for a given circumstance, there will probably be a list of several "third-related" voicings that are "inside,"
and other "third-related" voicings that are "outside."
In the mean time, though, most if not all of the
non-third-related polychord voicings have no place where they fit
completely within a major key (without accidentals), and so you could argue that the more "inside" voicings are to be found among those that
do involve a third jump from the root. It's just that there are some "outside" voicings in that group, too. Clear as mud?
Hopefully that answers your question, but because you've got me talking about one of my favorite subjects, I'll keep blabbing, for whatever it is worth to you and/or others here.
Before talking about things like elevenths and thirteenths, let's just consider sevenths. Specifically, let's consider that you can think of a seventh chord as being made of some root with some triad located at the third. Here's the usual list:
Maj7 = Root + minor triad located a major third above, as in Emin/C
Dom7 = Root + diminished triad located a major third above, as in Edim/C
Min7 = Root + major triad located a minor third above, as in Eb/C
Half-dim7 = Root + minor triad located a minor third above, as in Ebmin/C
Fully dim7 = Root + dim triad located a minor third above, as in Ebdim/C
(If we really wanted to be thorough we could include every possible version, like maj7#5, min-maj7, etc.)
Now- the question of whether any of these is more or less diatonic is less a matter of the chord itself, and more a matter of where that chord fits into the key. So, for example- is a C7 (a.k.a. Edim/C), more "inside" or "outside"? Well, in the key of F major it is simply V7, which is completely inside. (So although this chord is generically thought of has having "one flatted note" in it, it still has no accidentals with respect to the key). However, if you're in Bb major, it's V7/V, which is less-diatonic. (One accidental). You could argue that in other keys, it might seem even more "outside," such as in the key of Ab minor, where this chord would most likely function as a tritone sub for III.
Likewise, whether or not we're talking about 7th chords, 9th chords, 11th- or 13th chords, it is again more a matter of considering whether those chord tones are diatonic to the key.
Forgive me if you already know all of this, but in the podcast I mentioned that there is a full "system" that jazz players use, of organizing complex chords in such a way that they can think of forming certain triads on top of certain roots or left-hand shapes. Many of these involve a triad located a third away from the root one way or another. However, because of what I said above, the real question of whether a third-related voicing in particular is more "inside" or "outside" comes down to that chord's place in the context of the key or key center.
As an example: a dom13(b9) chord can be thought of as the seventh chord on bottom, with a major triad located a minor third below the root. So there's a voicing like what we've been talking about. A G13(b9), then, would be E/G7. Now, is this "inside" or "outside"? My point is that it depends on the key, and on accidentals. In C major, this chord is "somewhat outside," because it has one accidental. In Bb major, though, the same G13(b9) chord now has three accidentals, and so it will likely sound more "outside."
To look at that one more way, you could consider the same chord being formed in different places in the same key. In the key of C major, G13(b9) sounds more "inside," while A13(b9) is more "outside." This is because G13(b9) = E/G7, and has one accidental in C, but A13(b9) = F#/A7, which contains three accidentals in C. Does that make sense?
For me, the synthesis of all this is to think in terms of roman numerals, and to notice what voicings yield which results, as in: over a V7/vi I like to use X polychord voicing, when I'm in the mood to be "inside," while I like to use Y polychord voicing when I want to be more "outside."
Sorry if the above was more than you cared to read, but hopefully I answered your question somewhere up there.
James
