I had once a discussion about skills and how to valuate them in miniature games with a friends of mine working at CNRS (National Research Center of France) as a mathematician (he have a PhD in complex matrices, to much complicated for me =)).
We agreed to say that as long as we generate new profils, it's nearly impossible to valuate very precisely the cost of a skill dependently of the base stats of the profils, cause we need to have an absolute overview of all what can happen. So we can't mathematicelly have a perfect point system as long as we still generate profil, which is often the case in miniature games.
by the way, with the case of only Basic stats / Point cost / Activation Points, this is a dice game, so we often think about it with the "average" point of view, meaning that we evaluate our chance of succes regarding the average result we would obtain. BUT, the "extrem" point of view is as valuable as the "average" one, we could think about the maximum succes each time we are doing an action, and this is increasing a lot the impact of basics stats. And with a d6 system like in Eden, the difference between having 4 or 5 in a stat is really important, and combined with other stats this can be huge.
For an example, look at the Bayakoyé, 50 points, it's the ultimate beast but it only have 4/7/5/5, compared to Carnifex 3/6/4/5 And Kandjara 4/6/4/5 => try it, i'm not sure that Matriarcy is the winner, even with good profils like those ones. Why ? because Having 7 or 6 in CBT make a difference and having 4 or 5 (+1 protection) in VIG is really important (SPD are the same and PSY doesn't really matter here).
It's important here because if Bayakoyé is capabale of doing once an extrem launch, it will kill the opponent, but the opponent can't kill it even with the best result possible, this is a huge difference.
A lot of other parameters can be took into account also, and espacialy special abilities
