I have recently subscribed to Strava and am now increasingly using the route planning (alternative so far was Komoot). I have learned that Strava uses the average of the last 4 weeks to estimate the time needed to complete a route. But in doing so, you partly underestimate tremendously. If, for example, the last 4 weeks mainly flat routes were driven and then you make mountain stages, then the time estimation approach is not true at all. These 2 methods would be much more accurate:
1) Using pace [kmh and VAM] specific to the segments which makes part of the planned route. For example, if a segment of similar length and incline has been ridden in the last 4-5 weeks, then it would be much more accurate to use the pace driven in this segments.
2) Segmentwise comparison with other drivers of same strength