After scratching my head for a while, I'll pose a question about vertical alignments, namely, why in today's practice is it common for the type image to be so large on the body that adequate linespacing necessitates additional leading?
Up until a point and at least through to the 17th century linespacing was built into the face. The practice was evidently standard enough. Vervliet for example, assuming all type is set solid, measures the span of several lines to determine size.
Naturally there are implications of legibility, particularly with regard to x-height. To that point, the proportion that matters most is the relationship of the extenders. As far as I can tell a large x-height which is also large on the body is really a matter of economy, at least in metal type. Even in cases cited for their innovation the result often falls short of elegance, as in Van Krimpen's Sheldon.
There are some advantages to using spaces native to a body in typographic layouts. Doing so means the margins scale along with the type, maintaining page proportion across sizes. Granted, indents and column gutters might easily enough be indexed to leading instead. At a certain point it feels a bit arbitrary.
I'd very much like to know what means one would today use in determining the image size on an Em. Is it at all common to optimize for a size in which the type might be set solid? I assume there must be some other consideration I've overlooked.
Also of interest, when did leading type become common and why?