When reviewing performance, alert, or event charts it would be useful if an admin could select a timeframe (like with the Zoom control) and then fill out tag information for an event that occurred.
For example, I look at the historical datastore metrics for the past 7 days. 5 days ago at 1:30 am the latency levels spike to a historic level. I would then use a Event Zoom control, highlight the the timeframe of the issue, and then be prompted to enter specific fields about the event (ie SAN Disk Controller cache failed. Issue resolved with assistance of SAN vendor). Then whenever I view charts I could selectively turn the user entered tags on/off if they apply to the object or turn the user entered tags on/off if they only apply to the timeframe, but not necessarily the object.
These tags will help self document known issues in the environment.
-MattG