I know the game is still in its earlier stages, however some sort of formation system would make combat so much better because it would..flow better? Work at a faster pace? I don't know about you guys, but i feel like having, at least for now, some sort of basic enemy attack formation/system. It's basically inefficient to sit there and watch my units stand in line to attack somebody. Not to mention if there's some sort of AoE damage associated with his weapon, then they all take damage they didn't need to.
Obviously you can order your units to walk around behind whoever they're attacking, but even then they slowly get back into a line - so it's quite a pain.
So, Here's my idea:
Let's say that that formation first off occurs naturally - you don't have to order it, it's just coded in the game. Imagine warcraft III's gold mine, and you're playing as the undead. Imagine that your squaddies in kenshi are those acolytes, and that the guy they're beating on is the gold mine.. so it would look like this:

[Formation for surrounding enemy units, enemy unit is the mine, squaddies are the acolytes]
So, you'd have let's say you have 4/5 slots filled with squadmembers on one target, leaving one "spot" open for another to join in the fight.
Of course in some scenarios, this setup wouldn't be ordeal.. however if there was some sort of radial trigger system that detected the number of aggressive opponents in an area, and how many swordsmen you have that aren't exactly fighting... then it would at least work. (3 enemy units detected, 15 "idle" swordsmen.. Relocate them and fill them into slots spaced out to be groups of 5.)
Now, this is of course something that you'd start with, or at least implement. Actually keeping such a system for a combat AI would be... primitive. However, this would definitely be an improvement over what we have right now.
Criticize it however you want, but honestly something like this would make the game a hell of a lot better paced, or at least we wouldn't have to actually wait for combat to end, but instead be entertained watching it.
Is this possible to implement soon? It's not really that important, as research and the like is a lot more interesting, but i feel like this wouldn't be too terribly difficult to code/wouldn't be too time consuming.
Just my two cents
Feedback, ideas? Concerns? Additions?