Regarding "stupid" AI.
I wonder if the trap we fall into is that we expect computer-controlled crews to "think" for themselves to some extent. They do not, and in SB the consequences become immediately apparent. AI TCs will do virtually nothing without direct orders, which they follow to the 0110 1100 0110 0101 0111 0100 0111 0100 0110 0101 0111 0010. Intiative is not part of the equation.
Yes, mission editor and planning-phase scripting are powerful tools in SB. But who has the prescience to consider every option even before contact witht he enemy? Once the #%&*$# hits the fan, it's micromanage or die.
Not necessarily, if you look at platoon level scenarios and use the planning phase to the full extent. For example, the LOS tool allows you to identify terrain undulations that will PROBABLY allow you concealed movement and/or hull-down positions. It is therefore a smart idea to line those ridges with battle position markers. You can even create a retreat route from each battle position marker, a (slow) assault route from there to an alternate firing position, a retreat route from there, and back to the original position.
Then right-click the first retreat route once, select "copy route chain", and paste it to other battle positions that you identified and adjust the pasted routes and waypoints.
You don't connect them with your platoons' "spider web" or routes yet. You just leave them unconnected in the plan. Then, whenever you need a platoon to be in the vicinity of one of those positions, you just connect that platoon with the first battleposition and they are safe enough to be left alone for a while even with artillery threat around.
So - no, you don't know in advance where exactly you'll need that platoon. But when you place enough of these constructs in the terrain, you are flexible and fast in your reactions, and make use of the computer-crews abilities to handle the situation without too much supervision.
If you drive into battle without a clear idea of what you want to achieve, when, where, and how, don't be surprised ... as they say, "The reason why a plan failed may be ... that there never was a plan to begin with."