Engle:
You ask: If one can divide solution requirements into functional and non-functional with a fairly clear divisional understanding of each, then why not the same for business and stakeholder requirements ?
You also ask: The DFD technique is generally used by tech folks and theres no mention of it in the Enterprise Analysis Chapter in BABOK. What if other techniques were to be used, how would this fit in with the non-partitioning model ?
The answer to your first question is in part given by your second question. The partioning of behavioral requirements unnecessarily into three types confuses BA's alot more than it helps them. It has lead to the false pigeon holing of data flow diagrams as being largely a tool to model the behavior or requirements encompased within an automation solution. That flat out wrong: The experts of the technique, such as Tom DeMarco make it clear that data flow diagrams are just as applicable for completely manual systems as they are for automated systems .
The DFD data store symbols, for example, are NOT meant to be just program or database files. The ARE meant to be simply data at rest, whether the data is at rest within a computer, on a piece of paper in the inbox on someones desk, within your wallet, or wherever. Data flow diagrams are just as applicable for each of the BABOK's three requirements types: business requirements, stakeholder requirements, and solution requirements. Such clearly illustrate s that there is no need for three types of requirements.
And by not knowing the real purpose of data flow diagrams - or use cases, or BPMN for that matter - this whole career field degenerates into "Ya gotta have the right feel", "Ya gotta just know via experience", and my favorite "Because the BABOK (or IBM) says so" rules when deciding which technique to use. This is very immature.
Tony