The transformation of genuine coordinates into integer values creates a virtual grid into which actual data is inserted. One could easily imagine that the distance between two successive internal values expressed in original map units gives the size of the grid (by default, vertical and horizontal grid sizes may not be equal, except for unbounded Long\Lat). This grid definition is used to define the internal precision of a given map.
Internal precision is of interest because it indicates the ability of the software to differentiate between points with different coordinates during input by assigning them different internal values, and during processing by keeping them separate and by showing their distinct coordinates. Original measures are rounded up to the closest integer value during transformation; "restituted" decimal coordinates look like the original only if the rounding of the decimal equivalent of the internal integer values reduces significantly the number of decimal digits. It is very possible that two points showing the same coordinates (for example, in an ObjectInfo window) are in fact distinct in the map internal representation.