Testing location data with edge cases, not just happy paths
An address in a well mapped city center tells you almost nothing about how your system handles a rural route, a disputed border, or a query near the poles. Test the hard cases deliberately.
Elevation data resolution: what "463 meters" really means
When an elevation lookup returns a number like 463 meters, it is tempting to read that as the exact height of the precise coordinate you submitted, measured to the meter. What it actually represents is closer to a sample from a grid, where each grid cell covers a defined area on the ground and the stored value describes the elevation somewhere within or representative of that cell, not necessarily your exact latitude and longitude to the centimeter.
This is what resolution, expressed in meters per sample, is telling you. A finer resolution means smaller grid cells and a value that more closely tracks your exact point. A coarser resolution means each sample covers more ground, and the true elevation at your specific coordinate could differ meaningfully from the stored value if the terrain within that cell is not flat. On a flat plain, a coarse grid and a fine grid will agree closely, because there is little variation for the resolution to miss. On a hillside, the same coarse grid can smooth over real changes in height that occur within a single cell.
This matters most for two kinds of use cases. If you are computing something like drainage direction, viewshed, or precise height above sea level for engineering purposes, resolution is critical and you should confirm the stated meters per sample against what your calculation actually needs. If you are doing something coarser, like flagging whether a region is generally mountainous or coastal, or estimating rough elevation for a large area, the resolution matters far less and a standard grid will serve you well.
It is also worth remembering that elevation values, wherever they come from, describe the surface height, which is not always the same as ground level in a strict sense. Depending on the underlying data, a value can reflect the top of tree canopy or a building rather than bare earth beneath it, particularly in dense urban or forested cells. Optional ground elevation data, when it separates the bare earth surface from what sits on top of it, is useful specifically when that distinction matters for your calculation, since not every use case needs it.
My Geocode's elevation lookup returns meters above sea level for a single point or a list of points, with the resolution stated so you know exactly how much ground each sample represents before you build logic on top of the number.