A good deal of many miniature wargaming rules deal with the flanks of units, how units can be flanked, modifiers for being flanked in combat, etc. Flanks were an important consideration for warfare in certain periods, thus it is only natural we see the issue of flanks somehow translated to the gaming table as well.
Flanks, together with things such as line-of-sight and the movement of units, are part of what we could call the "geometry" of the game. We play on a two-dimensional gaming surface, units occupy a certain footprint on this surface (often a rectangle), and the orientation and distances of these rectangles relative to each other are an important aspect of many wargaming rules.
This implies that rules need to specify exactly when a unit is flanked by another unit, or can engage another unit in the flank, whether for close combat or for ranged fire. This translates into definitions of the front, the flanks, and the rear of a unit, often shown as geometric diagrams. The examples below simply give a variety of definitions I have seen in rules over the years, and I am sure there are more.
- When being attacked by more than one unit in close combat, the unit has negative modifiers in combat.
This covers the idea that you can receive a single charge always to your front, but multiple charges will most likely be in one of your flanks. - When being shot at by more than one enemy unit, the unit has negative modifiers for determining casualties.
This is the same idea as above ... - When shooting at multiple targets, the shooting unit has negative modifiers for firing effects.
Again, the same idea, but in reverse. You can orient yourself easily towards a single target, but if you want to engage multiple targets, it comes at a cost.
I would call such rules "combat flanks" instead of "geometric flanks". Flanking is abstracted as a concept, and is translated to "engaging, or being engaged by, multiple units".
In some of my house rules, I have used such ideas, both for large scale engagements as for skirmish fights. They do not always translate into combat modifiers, but sometimes into "actions points spent" as well. E.g. in one of my skirmish games, figures have to spent action points to do anything. Instead of specifying firing arcs etc., I simply say "shooting at a different target costs you an additional action point", thereby abstracting the cost for turning or meeting a new enemy. It also has the advantage that as a player, you have to worry less about the micro-precision of positioning units during the movement phase.