What would the bonus to the player be for implementing something like this?

Quite simple. If you don't want them to merge give them different end points.

Units on March orders to a point on the far side of a city will attack units in the city but maintain their stack separate from a stack sent to the city

It takes less time to give them proper orders than to have another button to use.

Have them go to the same place for a battle, and then add target so they stop in a different place after.


Simple


I don't know how else to describe it.

This game is not a simulation nor an action game, but a simple board game where the game pieces (units) where always placed on the selected playing fields (waypoints) one above the other.
That is the basic principle and also the challenge of this game.

As Vorlon already explained, in the case of close combat this can be avoided by splitting units and give them different end-targets (waypoints) outside the merge point.

But that only works with attacking units (and logically it may only work with them), because each attacking unit / army then has its own attack though, but simultaneously always also receives it's own separate defensive fire from the attacked enemy unit / army at the merge point.
And of course this option has to be actively managed by the player.

If one were instead able to split and set up also standing (defending) game pieces side by side on the same playing field (waypoint), then an attacking unit / army could attack only one defender at that playing field (waypoint) but would receive defensive fire from all defenders there - so we would then have the same problem on the ground (or at sea), which currently exists similar with the planes - unless, you would set up the entire game / game mechanics new ..

In that moment you become also an attacker and then comes in effect this:


I rarely use direct attacks, but as far as I know will separate squadrons of planes with separate attack orders fly separate attacks and not merge, also against the same target, as long as they either get a new order or are destroyed themself or the target is destroyed and they finally land.

Creative use of the delay feature will accomplish what you want as well.

While the idea itself is neat, there are the following problems:

- As Restrisiko mentioned the combat system currently works in the favour of the defender if the defender has multiple large armies standing in the same location. We would need to rewrite the whole combat system to fix this before implementing this feature, as otherwise it would lead to gread imbalance.

- It would lead to some players abusing this and splitting for example 30 units into single units. The whole map would become totally cluttered with layers and layers of troops and by this visually unplayable. Would need to come up with a solution for that first.

Of course, that would be perfectly correct for a simulation, let alone for an action game, but as stated above, CoW is merely a strategic "board game" with a corresponding game mechanism, and it's also a browser game where in most cases the opponents are not online at the same time.
So you have to set the microscope a bit sharper and take a closer look at what would happen if there was a change ..


For example, what solution do you suggest for the assumed case that a player with 100 tanks, for example, sends an attack order to an enemy position with as many as 100 tanks, and shortly before the battle begins, the defender divides his 100 tanks into 4 groups of 25 (or 3x 30 and 1x 10, or whatever he wants ..), so that there are standing 4 defending groups side by side on the same waypoint (and remember, all units make 1 shot each hour) ...


Without completely rewriting / rebuilding the game, I see alternative options:

1)
The attacker can attack only one of the four groups, so the 100 attacking tanks will attack only 25 tanks (or 30 or 10 ..), but receives defensive fire from all 100 defending tanks. Worse..
And which of the 4 groups does he attack?
Would that randomly chosen by the system? Should it be a matter of luck, despite an concrete attack command entered, what really will be attacked? Worse!
Or should allways the original group from which the others were split out be attacked? More worse..
..optimally the defender could then hold 1 tank in the original group, and on the first shot change this ONE defending tank would be attacked by 100 tanks, while the other 99 defending tanks could fire defensively without being attacked.
Both variants would bring a total chaotic distortion of the battle conditions, the game would no longer be strategy but a lottery.
So it whould be a complete nonsense to implement such an option here.

2)
The attacker attacks only one group, and receives only from this group defensive fire.
So, instead of attacking 100 defending tanks, the 100 attacking tanks can focus their firepower on only 25 (or 30, 10, or whatever ..) tanks, and only this 25 (or x..) tanks make defensive fire.
But which of the 4 groups is attacked and is the only defender of the waypoint?
Is it allways the original group, or is it chosen randomly?
Either way, with this solution, allways fewer units would defend than are available for defense - and which defender really wants to have something like that?

3)
The attacker attacks all 4 groups and receives defensive fire from all 4 groups.
That means, all 100 attacking tanks fight against all 100 defending tanks, and all 100 defending tanks fires back.
..but stop.., moment, that would be exactly that as it already is: allways all units which will be attacked will defend in the same moment.
So no reason to change anything.

Or according to which other option should, in your opinion, the thing work ..?


PS:
As already stated, the ability to split attacking armies into individual groups (by entering different endpoints after the combat point) is not the same as splitting standing / defending armies, as each attacking group fights separately and independently.
For example, if 100 tanks attacking 100 defending tanks, and the attacker splits to 4 groups of 25 units each (or 3x 30 and 1x 10, or whatever he wants ..), then each of this 4 groups separately attacks the 100 defending tanks, and all 100 defending tanks give allways defensive fire against each attacking group.
You see at first glance, the splitting of attacking troops does not even make sense in every case, but must be considered case by case in terms of unit number and SBDE.
BUT, even if the attacker can split his units, allways all units which will be attacked will defend in the same moment.

All listed examples for a better understanding without the inclusion of X-Factor, SBDE, terrain-, fortress- or home-bonus, season, weather, temperature and daytime, and also without consideration of units already destroyed at the first hit, since all that has no influence on the principle.

________________________________________________________________________________________________________
Dear reader, please, don't worry for my horrible english, i'm an old damn Kraut and learned it, while drinking beer with British soldiers in German pubs in the beginning seventies - so simply drink a beer and imagine you are in the pub as well, then you can read my writings without probs....
_________________________________________________________________________________________________________

That's why I wrote this:
Nobody really wants it that way.