Conclusion: Grid, or Ground, Depends on the Question
This companion opened with two points, a hundred meters apart, measured by two instruments that could not agree, and both correct anyway. Every chapter since has been an answer to the same underlying question, asked from a different direction each time: grid, or ground?
A GNSS-derived number and a ground measurement disagreeing is not, by itself, evidence that anything went wrong. It is often the expected result of two correct answers to two different questions, and this companion's entire job has been teaching how to tell which question is actually being asked.
Everything This Companion Actually Covered
Elevation alone separates a GNSS-derived distance from a ground distance, with no projection involved at all. A projection adds a second, separate distortion, grid scale factor, the moment a flat, stakeable distance is needed. The two multiply together into a single, practical combined scale factor, small on a short baseline and large enough to cost real rework on a half-mile corridor. Convergence bends direction the same way scale factor bends distance. A units mismatch, still an unfinished transition rather than a settled one, can silently offset a large coordinate value. Low-distortion projections and project-specific ground systems both exist to manage these effects directly, one formally documented, one often not. And an old deed, measured decades before any of this existed, was never going to match a modern GPS survey for reasons that are entirely technical, whatever a licensed surveyor ultimately decides those reasons mean for an actual boundary.
None of these causes are exotic. Together they explain the overwhelming majority of times a GNSS-derived number and a ground measurement disagree, and every one of them is checkable in minutes once it is known to look for.
The Question Worth Carrying Forward
Grid, or ground, is not a question with a single correct answer waiting to be memorized. It is a habit, the same way this series has built a habit with every companion before this one: before trusting a discrepancy as an error, ask what question each number was actually answering.
A GNSS distance and a ground distance are not guaranteed to match, and that is not a flaw in either measurement. Elevation, projection, units, and convergence each explain a specific, recognizable piece of why, and asking grid, or ground, at the first sign of a mismatch is what turns a confusing discrepancy into an explainable one.
The next unexpected discrepancy this method gets applied to will very likely turn out to be neither instrument's fault at all, just two correct answers to a question that was never as simple as one number.