I like to translate various technical diagnostic measures into similar organizational diagnostics, for instance:
Bluehatseo has a good post mentioning the percentage of a website's pages that are actually indexed by search engines.
This got me thinking about the optimal percentage of project-types done at Judy's Book. The majority of work/projects done should concretely move the business forward either by noticeably delivering more value to a user, or bringing in more revenue. Additionally, there needs to be an appropriate amount of maintenance and less-visible improvement work done as things tend to degrade over time if left alone.
The low hanging fruit here are projects that do not fit into either of these 2 boxes - and there are always some that don't - we can de-prioritize these. It helps to consider what your site is fundamentally about - for instance, if your site is about putting great content in front of visitors, then adding and improving content is much more important than displaying existing or sub-par content in a *new* way.