Distinguish Discovery and Data Model information based on circumstances

[csarven]

The naming of these sections and what's detailed should closely use/resemble on the list in the introduction section about circumstances in which profiles are used/shared.

Here are some general comments:

Perhaps rename section "Discovering a complete Solid Profile" to "Profile Discovery" or just "Discovery".

The Discovery section should start off with how a WebID may be discovered. It doesn't necessarily need to specify normatively because anything could virtually link to a WebID, and that the actual starting point of this specification is about the details once a WebID is discovered. So, factor in something along the lines of the following early on: "Anything may link to a WebID... Choosing the most appropriate resource from which to begin discovery of a WebID is at the discretion of the sender or consumer application." (But use terms from Terminology and Classes of Products).

The Discovery section should mention detail what "loading" a profile entails. It can remind/refer to Solid Protocol's reading resources. So, it doesn't need to specify any new protocol level requirement here, unless of course that's desired.

I'm not sure whether a Data Model section is needed because the specification currently has flat sections and that's fine. Perhaps rename "Extended Profile Documents" to "Extended Profiles".

"Other predicates" should perhaps be "Additional Descriptions" but it should actually just break down to specific purpose.