At one time, I was worried that the Clobberers were too weak, and considered exchanging the WA and FAD; this would give a starting position of (BD), (FAD), (WA), (NB), K, (WA), (FAD), (BD).
In this starting position, there is still a tangle at e3/d3, but at least the WA can develop via a3/h3 or c2/f2 if the central square needs to be occupied by a Pawn or a FAD. On the other hand, the FAD is now limited to b3/d3 or g3/e3 for its natural development, and this does not seem good.
More recently, I have wondered about starting the WA on h1/a1; true, this gives up the speedy development of the BD (by moving b2-b3, for example), and it also makes the plan of development rather predictable (your opponent *knows* you will develop with (WA)a8-c6), but it does relieve the tangle at e3/d3 somewhat. I think that (WA), (FAD), (BD) may be better than (WA), (BD), FAD).
This is all just speculation, with no real answers. It is possible that the Clobberers are already exactly as strong as they should be, and that a different opening position, if it turned out to be stronger, would make them too strong.
It is also possible that the Clobberers are so weak that only by making some of their pieces stronger could they be helped. The chapter in this document about endgame values versus current values was written very recently, and the impression of its arguments is so fresh in my mind that I might now tend to underestimate the worth of the BD and FAD...
Without going into excessive detail, I would guess that the positions where the Rookies start with the HFD on g1, or the R4 on g1, are inferior. Therefore only two starting positions are worthwhile.
I suspect that the standard opening position is the best.
So who makes retreating moves in the opening?
It's interesting that some of the armies have clearly-best opening arrangements, but others do not.
It would be too crazy to try to figure out if some armies should have different arrangements depending on their opposition.