The only thing that needs to happen to get an accurate reading with an incident meter is to "be in the same light as the subject matter". The easiest reading to take following this rule of thumb is the one where the meter is pointed straight back at the camera, because no offset is required to find the camera setting, you just use the numbers the meter spits out.
That's not the only way to use an incident meter. If you per chance are in the shadow of a hill where you can't see the main light (sun), you can many times still take the reading the same way you just apply an offset, in this case probably about 2 stops.
In practice, in the real world, this idea the equivalent of pointing a spot meter at a specific point in the scene and then deciding how that reading relates to the scene.
With very little practice the relationships/offsets of alternative incident readings are pretty easy to figure out, consistent, and reliable.
Once you understand and practice this concept there are few if any situations that would truly keep you from making a usable and accurate reading with your current meter.