A technically excellent mapping system can still sit unused. The difference usually lies in non-technical factors.
We once delivered a system that met every technical specification yet was opened only a handful of times a month. That taught us the success of a mapping system depends far more on how well it fits the user's workflow.
Start from daily work, not from the map
The first question should not be which layers to display, but which decisions the system is meant to support.
Remove steps rather than add features
Piling on features pushes users back to the old way of working. A good system reduces the number of clicks for the tasks done most often.
Design the update path from day one
Data that is never refreshed gets abandoned, so the update mechanism has to be designed alongside the interface.