Art doesn't fit in a box.
During the development of the requirements, I kept a mental note of how I thought the design would look considering the current requirement set. I had originally envisioned something like a strictly defined chassis with a backplane that the module cards slide in to with the face plate and any controls on the card itself. This idea is great from a standards point of view, because everyone has the same "box" to fit in to. The downside is that it severely limits who can actually build one of these pedals. The average guitar player doesn't even know what a 3d printer is, so unless this chassis is available in every music store in america, it's a bad way to start. This also forces anyone who is out to create an artistic pedal to be limited to choices of paint.
Art can't be confined to a plate.
Freeing the design from the confines of a rack style chassis, my next thought was to flip the insides over, and have a backplane that mounts to the underside of the top of a box, and have the modules plug in from the bottom. This frees the top of the box to be any shape and size that the artist wants... except it can't be smaller than the backplane. But what if I just want a buffer? I don't want a giant pedal for one module. So how big does the backplane need to be? Four Modules? Eight? What about a backplane that is perforated and can snap apart? This is getting complicated...
Keep it simple, stupid.
Somewhere in all my thoughts on backplanes I realized that if someone has to do something more complicated than plug a new module in, it's to complicated for most of the target users. Then I remembered that there is a bus technology that has recently become wide spread, requires no backplane, and expands with every module added... the "shield"! Brilliant! Using the shield bus concept, I can stack my modules, and completely break the dependence of the enclosure shape and size on the internal electronics.