When I view a device that has an agent installed, the CPU and memory graph has a nice control that lets me see the exact time if CPU or memory usage as I mouse-over the graph. I can also choose the time scale - 24hrs, 1 week, 1 month.

When I view a device that does not have an agent (like HTTP, TCP, or Generic), I have a similar graph, but it does not let me choose the time scale nor does it let me see the exact time for an event on the graph as I mouse-over it.

Seeing as how I monitor the internet gateway devices at each client, it would be very useful to have this extra information. It helps determine the exact point when a service went down or came back up. If a service is up and down, it would be really helpful to be able to zoom in and see the up and down behavior in more granular detail. Additionally, the graph only shows the availability of the device. It would be nice to see different lines in the graph for different monitored ports (in the case of a TCP device).