My flag usage has varied over time. I'm a big fan of start dates, but many items will pass their start dates (or in the case of a sequential project, not have a start date) without being started, and thus my list of actions which could be worked is pretty lengthy. I have a "tickler" perspective that shows actions grouped by start date which provides a steady stream of new work inputs, many of which will get done in the day or two after they first appear. As I review the older material, either by looking backward in the tickler's field of view, or as I review projects, I'll flag items that I want to get done in the next few days, and projects that should get the bulk of my attention. Every morning I'll look at the items due that day, the items that start that day, and try to finish all of the former and many of the latter, then attempt to empty the flagged list. I'm experimenting with a forced purging of the flagged list by dragging the whole list to OmniOutliner at the start of the week, then checking back at the end of the week to see if there are items I didn't complete. Theory is anything that I didn't do after a week of being on the list probably either needs some refinement, reconsideration, or maybe I just put too much on the list.