>>"pj; this seems to be right up your alley.. ever thought about pack density?"
Can't say that i think much, Brett, 'bout anything that is.
Well, ok, sometimes. Whether they're good thoughts or not, i'm not always sure. They seem to be at the time. I'm working this out with PacerPoles on steep slopes right 'bout now, and don't know if i can spare any CPU horsepower to simultaneously work on another problem. But, here goes...
i try for a dense pack. regardless of the pack. w/some UL packs, compression is lacking (G5 and G6 Whisper for example), so there's only so much one can do using the innate compression capabilities of the pack. i figure, rightly or wrongly, that dense packs have lower rotational moments. Does that make sense to you? Maybe i'm not expressing it right. I'm just a lowly computer/software engineer, with only a rudimentary understanding of all things mechanical and physical.
Brett, regarding your comments on bending, etc. In my old posts on CG and packing, i've mentioned this very thing. I'm often bending and twisting on some trails, or stepping over and around roots and rocks, and deadfall. I find my unique triangular loading scheme works pretty well. Sure, it involves tradeoffs, but unless one is just going straight line nearly all of the time, i find my scheme works well even if 99+% of trekkers pack conventionally (i.e., hvy-hi) and disagree with me. I'm used to being strange and different (one look in the mirror each morning confirms that assessment).
As far as the bodies natural unladen CG, i'd modify your navel comment slightly, and move the CG half-way back (front to back, that is) into the body (that way it doesn't matter if one has an "outie" or an "innie" :).
Also, to simplify, i view this as a 2D problem and NOT a 3D problem. Why? No one i know disagrees that we should attempt to distribute wt evenly left-to-right. So, i "factor" it out of the equation. The issues are front-to-back and low-to-high; how to optimize that for the type of trek we are going on and the expected pack weight. Climbing (which i really know nothing about), i'm guessing (someone who climbs, please correct me if i'm wrong) keeping it as close to the back as possible is probably good to reduce the rearward moment arm that would tend to pull one away from the face being climbed. Same would go for a very heavy pack, but perhaps not as extreme, just enough to reduce rearward pull on the shoulder straps. L/UL pack weights, unless one is just going straight ahead (and then reducing rearward pull as much as possible might be best) with no bending, leaning, or twisting, i'd say my unique triangular heavy low and nearest the back is best. It feels good when bending, leaning, and twisting.
Your explanation about the vertical moment coming into play when bending, etc, is precisely my reason for packing as i do. Theoretically, it simply makes sense; it's logical and to me intuitive (though perhaps b/c of childhood experiences). Practically speaking, it just plain feels good.
Some of the old posts dealt w/male vs. female CG, and bodybuilder type physique CG vs Sumo CG.
Best CG possible - make a "Sumo Pack" (y'all can stop laughing now; i can hear y'all way out here in New England; don't go gettin' Gimli riled!!) - weight distributed in various pockets around the mid-section just like a wrestler (at least water bladder[s] on side and in front). Obvious limitations include being able to see one's feet for proper foot placement, non-interference with the thighs when stepping up high, and keeping one's body close to rock faces when scrambling or climbing. The Sumo Pack would also have some small volume on the sides of the legs (not too much b/c this makes the legs too wide for squeezing through tight spots; very handy for fetching things easily though), thus distributing some of the wt below the body's natural CG to offset that weight carried above.
I know, some crazy packing ideas.