The Case for Squares

One of the issues with hexagons is that they tend to be the wrong size. Now, of course you can choose to have smaller or larger hexagons, but the thing is that no matter which size you ultimately settle on it's not optimal for all purposes. Part of this is that the natural unit of a map is distance, while the natural unit of travel is time and the unit of exploration is line-of-sight. Thus, plain-hexes could reasonably be much larger than jungle-hexes.

Another part is encounter density. To more accurately represent your setting, the solution is to zoom in. But once you do that you get a crazy number of hexes to fill which a) renders you a massive amount of work; b) tempts you to cheat the players of their agency by having totally random contents (like MYZ); and/or c) creates a strange pace when players by chance repeatedly choose the unpopulated hexes.

A basic cause for these problems, I would argue, is that space gamified through hexagons conflates unit of movement (ie. "move one hex") with points of interest (ie. "in this hex, there is..."). Or in other words, hexes scale poorly. There are solutions for dividing a hex into smaller sub-hexes, but they aren't very practical.

On the basis of this, I want to suggest another way: to treat the game-map as if it was a normal map. I have yet to playtest this, so there might be some hidden flaw, but since rules are often more complicated than relying on prior knowledge I think it might carry some merit.

The idea is this
Instead of superimposing hexagons on your map, you place a square grid.
This square grid does not primarily represent spaces in a game sense (as in: move your pawn two spaces) but a coordinate system.
Make the scale so that 1 square is 1 (metric) mile square in reality, but 1 cm2 on the map.
Set base movement to 3 miles in light terrain.
Use a ruler.

This has the following consequences
For MOVEMENT. The party can move 3 squares horizontally or vertically, OR that they can move 3 cm in any direction they want. So a novice group would probably stick to the orthogonal, whereas a more experienced group would say "we travel NNE for eight miles, then turn sharp E for another seven to stay clear of the Dragon valley"

for SCALE. You could zoom in and out as you like. Each square can be divided into 10x10 new squares (each 1 km) or 100X100 (each 100 m), and each 100 squares in turn form a macro square (10 x 10 miles) and so on. And as long as you keep this basic structure, your coordinates will work. As an example, East 20,0 through East 20,9 would be the ten horizontal locations nested within square East 20. And East 20 through 29 would in turn be part of the macro-hex East 2X.

for ENCOUNTERS. Since the location of anything on the map can be stated as precise as your ruler allows, you could vary the encounter area for each encounter. A hidden cache is perhaps only found is the party moves into the exact dot representing it on the map, whereas a city is encountered in the square and a prowling beast in encountered within a four mile radius of its lair.

for LEGIBILITY. Normal coordinates are much easier to track than hex-coordinates. There's no need for numbers in every square - making every fifth gridline thicker and having numbers every tenth would suffice. This allows much more room for at-a-glance information on the referee's map.

Obviously, there are strong reasons for sticking with hexagons. They look good. They rest on a strong tradition, and form the basis for many procedures. They have an esoteric quality that sets game-specific knowledge apart from normal knowledge. Etc. On the other hand, I think that there's a solid case for squares.