Yah, list bloat is a problem I have too. I think my biggest weakness is in the area of projects and the weekly review. Basically, when I first plan out a project I brainstorm what I think all the next actions for the project will be. That's fine as a way to start, but the problem is that I neglect to prune the project list as those next actions change, so I end up with a lot of defunct next actions and a headache of a time trying to reconcile my projects with my next action lists. Sometimes, too much information can make it difficult to make decisions as to the next action, so I've tried to go with a more minimalist approach, placing only 1-3 next actions on each project list, which forces me to make decisions at the right time--i.e., when I have a clearer picture of what they should be.

I've also started using kinkless GTD, which is a lifesaver, since it offers a painless way to remove items from my contextual lists and to reorganize the N/A's within projects. In this sense, atomic next actions that can be reshuffled offers an escape from the limitations of static lists. The same effect can be achieved by using a pack of index cards as a list (one action per card).

As an aside, I suppose this is why in the olden days of GTD there were so many debates about whether projects were allowed to have more than one next action. As far as I can tell, these debates were settled decisively in favor of multiple N/A's per project, since there are usually several N/A's in a project that are not dependent on one another. While I agree that a Next Action is "anything that can be done now," the concept of one N/A per project nonetheless prevented lists from being flooded with information.