In the case of the loopback interface, since it typically doesn't lead to any significant performance degradation, you can exclude the interface. To do this, you would open Diagnostic Server Administration, select the appropriate (or default) connection, choose the Network - Percentage Bandwidth alarm, and then "Override the default settings..." to "Add..." the MS TCP Loopback interface and then change or disable the alarm severities and notifications.
Spotlight's got a default threshold set for network interface saturation, just as it's got default thresholds set for all of the alarms in the product. The approach is of a general nature, meaning that when a DBA installs the product, we will alarm on anything that appears to be amiss and allow you to determine whether action should be taken. The Alarm editing capabilities have been designed to allow you to configure exceptions and/or specific actions down to the lowest level of granularity available to ensure that you can customize the product to work as your infrastructure requires. In some cases, like with the Days Since Last Backup alarm for example, there are clearly databases that should be ignored so we do that by default. When we start to delve into things like network configurations however, individual customer sites have unique implementations that make default exclusions inappropriate or dangerous, so we err on the side of caution in those cases and allow you to control how the product responds in your environment.
I hope that helps