Feedback on tagging in Silver IA prototype

This is further to comments I made last week about how tags were and
weren't used in the prototype.
I wanted to add a further point, which I didn't emphasise enough, about the
potential problem with a purely "flat" architecture in general.
Currently, if I haven't completely misunderstood things, it seems the
suggestion is that the guidelines will be presented without any overarching
structure or organisation. There would simply be a list of 78 guidelines as
the default view.
Many different people need to get an understanding of the guidelines in
their entirety, and some kind of presentational organising categories for a
default view of all the guidelines really helps that.Without  them you
simply have a long list.
 This doesn't mean these categories have to any normative role, they would
simply act  act as organisational devices to make the material more easily
understandable.
If there isn't any structure, what is the intended default view of all the
guidelines? (The same goes for methods)
Mark Tanner