I've also put up a quick jpg of the {c} and stroke based sets, http://voynich.info/vgbt/strokes.jpg. More glyphs will be added to these sets as pages are transcribed, but it becomes fairly obvious (to me anyway) that there is a logical and predictable pattern to their construction, and that what holds true for one glyph form tends to be true for other glyph forms.
I have my own theories as to why these things are drawn predominantly in sets of four, but Roman numerals also works as a theory! ;-)