"only XL-2"

The idea of using random, anonymous templates is actually good, if it weren't for the awful blue color on 6LM10A and 8HM12A.
Rebellius, could you tell me if there's a way to prevent color imbalances on these templates?

By the way, on the last LM, I got a Necropolis hero with a good starting location on muddy terrain, but that doesn't happen often.
And why do we always play as red and blue? There are other colors too. There's no statistics on how skewed their distribution is.
By the way, have you noticed that no matter how big the map is, blue and red are most often generated on adjacent starting locations.
Even the simplest Jebus map always has surprises in terms of the contents of the starting locations. You could, of course, always set a high density of object distribution, but that would be boring.
From the theory of zone generation:
Zones treated as circles at this stage
Important concept at this stage is zone’s center
Placement done on arbitrary grid, translated and scaled later to actual map size
First zone placed in center
Later zones placed adjacent to existing zones
Zones allowed a small amount of overlap
Minimize overall width and height of map
For two level maps, each level must have at least one zone
Maximize number of adjacencies to connected zones
Zone placement done 5 times, like “shaking a box”
Extra water and rock zones placed after normal zones
Voronoi diagrams are responsible for generating the zones (distribution of zones on the surface). Somewhere in the generator, there's a formula for this, but who will find it and correct it properly, and is it even possible?