1. The most obvious problem with the diagram was its lack of flexibility. If I took a wrong turn at some point, I had no diagrammatic way of a) finding my way back to the route or b) recognizing the route when I did get back to it. There could also have been other problems like ambiguous or non-unique landmarks. On a more basic level, the analogy between the physical route and the diagrammed sequence could have been faulty. Any of these problems in application might have led me to seek a different or more comprehensive diagram.
2. The diagram itself might have been inconsistent. This is not too common with something as straight-forward as a map, but it is all too common with more complex diagrams relying on things like logic or authority for their internal consistency — not to mention ideologies which are often intentionally inconsistent.
3. The diagram could have produced consequences that didn't play out. If it didn't lead to the proper destination, that would be a biggie, but there could have also been lesser inferences, things like indicating a left turn where none could be made or not clearly indicating which road where more than one veered off in a direction, that would have called for some kind of revision or notes to the diagram.