There's no actual science behind this, but:
I've always thought that stable jump nodes are relatively stationary. They can drift a bit, probably due to their interaction with local gravitational fields and the variation of position between the two star systems (not to mention any unknown interstellar bodies or dark matter that might be lying in deep space). So in the course of a few hours there might be no noticeable movement--but over several months they might drift as much as dozens of kilometers, mere peanuts for interstellar travel. This would give a reasonable excuse for ships jumping out of position, and having to physically travel to the nearest node.
Regular navigational updates would be uploaded onto FleetNet or some civil network on a regular basis.
As I was writing this, another thought occurred to me--there might be some kind of dangerous interaction if a ship were to jump into real space too near a jump node, especially if the local conditions haven't been evaluated for a long time. Like say, accidentally juming to the next star system--or worse being thrown way off course and into an unknown system. (Or worse.) That alone would make a good reason not to jump too close to a node.