I'll be honest; I don't think you can quantify balancing in such a simple way. It varies hugely with a number of factors that are not easily quantifiable such as ease of use, micromanagement, player mentality and role. A quantified method such as this would only make every unit equal to every other to some degree, which in my eyes is rarely justifiable in an interesting game design because certain units should be weaker by nature but required or useful for certain tasks.
The engineer in RA2 would be a good example of this; how do you determine the value of an engineer? Really, the only way is playtesting.
Balancing isnt decided by any formula, its all through testing and recoding small parts to get it to work, its like the nuke cannons in Zero Hour, while its weapon is really powerful, you have the problem of setting it up before you can fire, those kind of things cant be accounted for by any formula.
Cost, Speed, Turret Rotation, Turning Speed, and just about every other thing.
And at a point, balancing doesnt mean they all have to be exactly the same, look at the GI in RA2, a deployed GI, costing 200 can easily take 2 conscripts costing 200, but that is balanced by the soviets having the tesla trooper to back up squads.
thanx for info! i knew that i missed some stuff of cost, speed ect.. but thanx for the backup squad info (the last part of ure post). I think i use this in my mod. now i have balanced it one on one but using squads its more interesting.