Every data network engineer has had to sometimes do a quick and dirty estimate of some network dynamic or factor. If you have an application with a message of size X, and so many people reading so many messages per unit time, what will that do to your link utilization? How much of an overhead factor is the protocol? How much delay can we expect?
The estimators are based on commonly used theoretical algorithms and will only provide estimates. Your actual WAN may or may not compare well to these theoretical answers. But they should compare favorably to those "back-of-the-envelope" results.
