I think we should be careful not to treat reports and books as the same thing. Ahnentafel reports have been used for hundreds of years by genealogists to convey information about pedigrees in a *standardized* way where numbers express relationships relative to the proband. The ancestor reports created by FTM and comparable programs are in this same vein. Although they may serve as a component that could be used in a book or even an outline by which a book could be organized, at the end of the day a book is far more complicated, and it should be treated separately and in more depth.
That is not to say that I don't think that FTM users wouldn't benefit from more choices in the way in which ancestor *reports* are organized. A different, more expressive numbering scheme is the next logical enhancement, but the alphabetical listing of ancestral lines you've suggested here and elsewhere also has merit, though to a somewhat lesser degree as it presupposes the use of family names and is more ideally suited for unique and unchanging family names at that. Still, it is a logical option that would no doubt be useful to many.
You are correct to criticize the binary tree model (which both the Ahnentafel and Ancestral Lines numbering schemes convey) for its inherent inability to express collapsed pedigrees. Binary tree models, as a workaround, must employ some type of marker or tag to denote an ancestor with one or more relationships to the proband, and that's easily done by virtually all genealogy software today. A more accurate representation, however, would be a directed acyclic graph, which I don't think is supported by any genealogy softwaer, TMG included. (We have shot off on a highly esoteric tangent.)http://en.wikipedia.org/wiki/Directed_acyclic_graph
(edit: added silverfox as my direct address since kathymarieann posted before I could get my reply out)