I'm sure there are pros and cons to each way, but I don't think we've really touched the surface of how each way differs from the other. For me, the best way to start is to compare a string for a so called symmetrical bow, ie. arrow pass 2" above center on a 4" handle, to one where the arrow pass is closer to dimensional center. I think what you'll find is the the upper limb is shortened in the sym bow as indicated by the nock point on the string. Not so much so on the so called asym bow. Makes one wonder what the term sym is supposed to mean, when applied to such a design as the arrow pass 2" above center.
From my pov, the nock point is the point where the bow will apply forces, the central point of the bow. I can see no compelling reason to put it so much above center as to handicap the upper limb, and give the lower a free ride. Particularly in a bow where we want to work as close to the elastic capacity of the material as possible, like with an all wood bow. For a glass bow, theres so much extra work capacity I'm not sure it matters as much in that context. But I think even in a glass bow having the string segments, and consequently the corresponding bow/limb segments be as similar as possible (the triangles these elements strike) makes it a heck of a lot easier (for me) to assure they both draw and return in relative unison. When they are significantly different shapes (the triangle the arrow, limb, string make) it's harder (again based on my fevered way of thinking) to bring them into balance with spring force (tiller) alone.
Here's a site I cite frequently for some interesting background reading. Not for the faint of heart. In fact, I'd wager the average bowyer would be better served to simply do what they've been taught or have come to prefer with trial and error, or experience.
http://www.goarchers.org.uk/mechanics/ In particular the section in [Bow Mechanics] titled [Tiller] discusses the asymetry inherent in the bow/arrow/archer model, and the corresponding implications for bow design and tiller. To boil it down, both positive tiller and placing arrow pass closer to center have the same result, they mitigate the tendancy for the bow to rotate upper limb toward archer on the draw stroke, and the tendancy for the arrow nock to travel in an upward or downward curve on the return or power stroke stroke, versus traveling in a straight line.
This is an interesting topic. One where I haven't found a really, really compelling justification for one option over the other in practice, honestly. And one where a wide variety of opinions exist and imho, a fairly shallow understanding of the underlying geometry and physics, myself included. As such, sometimes folks can get a little testy during a long, detailed discussion. So while it interests me, I try to tread lightly in order to honor other folks preferences and pov. But I make most of my bows with the lower limb 1/2 to 2" "shorter" (which is a misnomer) than the upper.
I've never quite accepted the notion that the natural balance point of the bow, if suspended horizontally, has any relevance to how it shoots, "handles" or how convenient it is to carry in the woods. It's part of Dean's argument favoring the arrow pass at center, which places the balance point closer to dimensional center, the so called "draggin anchor" argument. I don't doubt it, it just doesn't come up as an issue for me when I'm totin' a bow around.
Everybody who knew I'd come along and write a book about this topic raise yer hands. LOL