Well, lets start with this frightening 1% that you seem to be so concerned with that makes the Brood vastly over powered. Yes, they can be killed 9% of the time, but they have a 10% chance to kill their opponent. I know, that vicious 1% is a game changer, but again you forget that Lash Out only triggers if they survive, which means that that 10% chance to kill their opponent only activates 91% of the time (since they died the other 9%). This means that they only have a 9.1% chance of eliminating their opponent, so now we are only looking at them being over powered "head to head" 0.1% of the time. Does this make them so vastly over powered that they are broken?
You asked "What special effect would I get on comparing models/number crunching by playing down multiple games?". As an answer to my question about have you even put any models on the table pretty much answers no. The problem with what you are doing is that you are trying to balance things on a component level. Which while possible, the only way to do it is to make each of the components identical (which would give us the game of chess, or go, or Pente or Hive or any number of simple games in existence where both players have identical units/models/tokens). To add flavor, and so as not to duplicate what is already out there, miniature games create factions and each faction tends to have their strengths and weaknesses, so you balance these models not on a component basis, but on a system basis. How things work in a vacuum of 1 is often quite different then how they work when they are grouped together. Your comment about flipping a coin in a single isolated incident, and not a system, so raw percentages work, but what if you were flipping a coin over the surface of a viscous liquid that sometimes allowed the coin to land flat, but if the coin hit is at too great of an angle it would slice into the liquid and get stuck? In this case, yes I would ask you if you ever bothered to flip the coin. Add to thus an opponent that could vary the viscosity at a whim, now how does your 50% chance of getting tails work out? The reason I as if you ever bothered to play is because then you get a better understanding of the game on a system level, rather then always looking at it from just ONE aspect of the component level. You examples are always worst case for the piece you want to lose and best case for the one you want to show as over powered. Every faction has strengths and weaknesses, why would I always attack my opponent in ways that favor them? Why would I put one Skorza vs one Brood. Wolves hunt in packs, what are the odds the Brood dies when 2 wolves attack at once? And while you show disdain for stalker and skirmisher, these allows the wolves to always engage the Brood favorably, so they will always be able to focus their strength in ways to negate the broods strengths. As for the Wrath, they have enough varied benefits from enlightenments, that a discussion on ways they can be used to face different units would take a novel to cover.
As for your solution to the Aadanii issue, will it actually fix what you find so abusive? Would it be too much of a nerf? You like to forget that the Aadanii are the only 1 rate Rank 2 infantry out there. your point of "You mean how the Skorza's ability relies on it to actually kill a model or how the Wrath's ability relies on another model and to be in that model's aura radiance? Your right, neither of these rank 2 infantry abilities rely on some outside factor" just supports what I am saying. When a Skorza attacks is completely in control of the Skorza player, if a Wrath is in an enlightenment aura is completely in control of the Shael Han player. When a Brood gets to use its Lash Out ability, that is completely in control of the Aadanii's opponent, since the Aadanii player cannot force the opponent to attack their brood. By outside factor, I meant something that the player has no control off. Aadanii abilities rely on the opponents actions instead of their own.
As for Hadross vs Shael Han - only looking at the rank 1 infantry (since Hadross Rank 2 is very tanky), and with how resonance has become much easier to spread we are again only looking at a 1% difference. A resonating Legionnaire has a 16% chance to kill a lone Deepman, while a Deepman has only a 15% chance to kill a Legionnaire standing next to another Dragon Guard, while under the Keepers Aura. But hey, since we are using leaders, the Deepman has a 30% chance to kill the Guy the initial target was standing next too if he manages to hit that 15% mark the first time.
I guess if you get anything from this, it is the point that you cannot just look at individual components to make balance judgements. They are good to get you in the ballpark, but you need to look at the system as a whole to be able to fine tune the pieces. You also cannot look at best case / worst case to determine balance, you need to find what the typical is. Stressing about how over powered something is over what on paper is a 1% of the time issue seems a bit drastic.