I'm a casual player, so correct me where I'm going wrong.
It seems like a lot of people think supply lines costs (15% per army) favor building nothing but doomstacks. People don't like this, and I can see why it would suck to do nothing but doomstacks: variety is more fun and more immersive.
However, I started trying to do the math in my head. It didn't seem that bad. I mean, armies are expensive, but I wasn't seeing where Doomstacks are heavily
favored by maths. Not trusting my head, I got out excel and started really crunching.
The cost of all your forces is equal to:
base_unit_costs * (1 + (number_of_armies -1) * supply_line_multiplier).
Imagine, then, 5 doomstacks. Each army has a base upkeep of 6500. This would cost 52,000.
Imagine 8 basic armies. Not total meatshields, but stuff you get from low tier buildings. Base upkeep is about 3200 per army. That would cost 52,480--or basically the exact same amount as 5 doomstacks.
But what if the supply lines multiplier were 2% instead of 15%?
Well, your five doomstacks would cost 35,100.
How many of the 3,200 basic armies could you field for that? About 9. So you get one more basic army moving from a 15 to a 2% supply line multiplier.
I find it hard to call that "heavily favoring doomstacks." Sure, the armies are waaaay cheaper, but for the same price, you get in the ballpark of the same doomstack:basic army ratio.
Now, if we move to say 16 armies--that's more than I've ever had before but again, I'm a casual player--then at a 15% multiplier, you would pay 166,400 for those basic armies. That same amount of money would field 10 and a half doomstacks. That's basically the same ratio as before.
If we had a 2% multiplier, you'd pay 66,560 for 16 basic armies. For that same price you could have 9 doomstacks (cost = 67,860).
Again, I find it hard to call that "heavily favoring doomstacks."
In case you were wondering what happens if we try to field total meatshield (Clanrats and Skaven slaves, about 1200 eacg base cost) armies, you could field 16 at 15% supply lines (=62,400) for the cost of 6 doomstacks (=68,250).
In short, the 15% multiplier seems to heavily favor fewer armies, but the ratio of doomstacks to basic units that one can support has not been greatly changed.
Here's an excel screen shot with all the data