The problem here, I think, grossly simplified, is something as follows. As a player you can say we will not turn phalanxes in ways that would be unfair, despite possible in-game. Whereas when you have several cavalry units with which you would like to attack an enemy unit, you cannot say I will tell my cavalry to attack in a long line, to cover the length of the enemy unit (as opposed to convergence). We simply cannot, because it is impossible in this particular game.

The example of the phalanx is one in which we are aware of a potential unfair maneuver and choose to avoid it.

But the example of the convergence of cavalry is one in which we are aware of the effects, but are at a complete loss. We have no alternative, it would seem. Unless, that is, you suggest the player use solely one cavalry unit to attack an enemy unit, in which case you completely cut the convergence route of more mass. Maybe he doesn't wish to use his cav elsewhere. Maybe his current goal is to break that unit of yours. What other option does he possibly have.

I'm not being articulate right now, but I feel like there's a categorical difference between these two examples. Let me know if I'm wrong.

Regarding the engine-blaming, as you put it, I don't want it to seem like an excuse so much as an acknowledgement. Our goal, I hope, is not to bend players too much in terms of regulating their in-game tactical choices, even by way of what we deem to be fair play actions.