I've always used the phrase pop-value density. I've only used cost as a rough index of value because that's what it is. While not perfectly accurate we'd spend years deriving some other index of value, and I know this for a fact because balance team has spent the past 5 years trying to perfect the already existing index of value i.e. cost!
You'd see this if you were honestly evaluating the merits of the idea, but it's clear you're straw-grasping instead of admitting to yourself that you rushed into an argument without actually thinking it through.
I fully agree on value density, however you have defined value density as purchase cost per pop over multiple posts. And with that I fully disagree.
Balance team has done a good job so far. You still have not shown in any way that your approach has any merits. Your only try was a theoretical Ober vs Penal comparison (based again solely on purchase cost) assuming they are already perfectly balanced pep population but not per purchase cost. But they likely are not, at least you have not shown so other than 'after 5 years of balancing I assume thiese units are balanced'. Which they are not, which is why there will be at least two more balance patches.
I have then asked you to explain how your ratio of purchase cost/pop (which was the only definition you gave gor value density up to the DPS post) can make predictions and how this value changes when you try to balance out units. I provided reasoning that it does not change with actual unit balance with both units as in the live game as well as a hypothetical scenario.
You have ignored responding to all of that and finally said that one needs to look at DPS. Which was, among other things, exactly what I argued for in the first place.
I am done with wasting my time on this.
Have a good one though.