As with all things photographic, we each decide where to point our tools and we provide the context needed to make decisions about what to do with the reading. It's the nut behind the camera that makes it work, not the tool in his or her hand.
Way back in time before incident meters got their domes a technique was developed to address the blown highlights problem on transparency film when shooting in high contrast situations, its called duplexing.
Classic duplexing takes two readings: one with the meter pointed "back at the lens", essentially the same as is taught in incident meter manuals today, this reading is to find the "best" exposure for the main subject matter; the other reading is taken with meter pointed directly at the main light source, sun or whatever, this is to find the "best" setting for the highlights. These readings are then averaged, just as is normally done with shadow and highlight readings from a spot meter. (This model makes certain assumptions, like "the mid-tones and the highlights are what we want to protect". If shadows are more important than say mid tones just replace that reading with a different orientation of the meter head.)
In both cases, averaged duplexed and averaged spot readings, a compromise is made. In high contrast lighting when using say Velvia, there is simply not enough film range to get all the highlights and all the shadow detail we might want. Something is going to be lost, both methods simply do their best to find the best balance.
Side note. After incident meters got their domes duplexing for front lit and cross lit scenes became almost unnecessary, the final camera setting is normally the same. Duplexing does still have real value though when the subject is backlit.