Here is the way I think of it, if a cylinder passed through two perpendicular planes, flatlanders living on those two planes, with different perspectives on life, might call each other to discuss their observations. One flatlander might report the observation of a circle, while the other might say they saw a rectangle. Both have observed part of the truth, but different parts of it, and to a flatlander it may seem these observations are irreconcilable. A third flatlander who decides to simply average the observations together into a rounded rectangle is more wrong than either of the others. The key to resolving the apparent contradiction is recognizing that the truth exists in a dimensional space higher than the flatlander is capable of perceiving.

Well what I'm claiming is not that averages can be used to arrive at all truth, or to observe all things. Many things however are distributions and averages can be accurately used to characterize them in a way that works. In that example, one might say that the mean viewpoint is correct within the error bars that encompass the two observations, which could be quite large if they are opposite observations.

Sounds like you two are saying nearly the exact same thing, using different words.