I believe these are very important points, and I'm grateful to be reminded of them and to be able to think them through. My impression is that it works like this:
We clarify every item of our "stuff" in the clarify step of the workflow process because we do not know exactly what each item of our "stuff' actually involves until we determine what it is. We do this thinking first, so we do not need to do it later. This makes our work easier because we have already done the thinking required.
Then, still in the clarify step of the workflow process, we ask ourselves whether this item is actionable. If it is not actionable, we make a note to file it appropriately as (1) trash, recycle, or shred (we have no use for it); or (2) Reference (we may need to refer to it, but no action is needed); or Someday-Maybe (we may or may not do something about this later, and meanwhile we are parking it for review). To keep clean edges between the clarify and organize steps of the workflow process, we do not interrupt the clarify step to file, so we do our filing during the organize step.
If the item is not trash, recycle, shred, Someday-Maybe, or Reference, then it is actionable. This means that there is something we need to DO about this item in order for it to be the way we want it to be.
If an item involves one action step only, it is called a Single Action. If the item involves more than one action step, it is called a Project. For ease of handling, unrelated Single Actions may be kept separate, or they may be associated together and treated as a Project. A Project is generally to be completed within one year. We need to bring both our Single Actions and the individual action steps on our Projects into our workflow.
*****
This is where I am not clear on the steps. In regard to taking action on our items, we need to ask ouselves, "What is my desired outcome?" Do we have a desired outcome for a single action, or for a Project, or for both? Do we determine the desired outcome before we determine the next action or afterward? I am thinking that I cannot decide on the next action until I know my desired outcome, so perhaps we now decide on our desired outcome. Let me know your views.
*****
Once we have determined that our item is actionable, and we have decided on our desired outcome, then we need to ask ourselves. "What is the Next Action on this item?" Then we state the next action in concrete terms as a simple physical action, so we will clearly know when we have completed it.
Only after we have answered the question "What is this?" and have answered the question, "Is this item actionable?" and have answered the question, "What is my desired outcome?, and have answered the question, "What is my next action on this item?" do we apply the two-minute rule to the Next Action that we have stated.
We ask ourselves "Will this next action take less than 2 minutes?" If we estimate that it will take less than 2 minutes, we do the next action now, rather than enter it into our GTD system because this is more efficient and faster. However, if the next action will take two minutes or more, then we put the next action on our Next Actions list under the appropriate context.
In my understanding, the two-minute rule is specific to the clarify step of GTD workflow processing at the specific point after we have already stated our Next Action and asked ourselves whether it will take less than 2 minutes to do it. It does not necessarily apply to any other situation. However, as far as I know, there is no reason that the two-minute rule cannot be applied whenever it works well for each individual, except that it could lead to lack of clarity, if the clarify process is not done. I recall that DA is very big on keeping clean edges.