Best at... what? Your whole scheme of balancing against IO's is based on the flawed assumption that there is only One True Build for each AT.
Data mining will give you information on the most popular builds. What happens to my scrapper (who is increasingly biased towards being a killing machine) if it turns out that living bricks are the most popular scrapper build? By default, if you're building around the most common, you're disadvantaging those folks who don't play/build around the most popular philosophy.
In the case of Brutes, there's a pretty good chance you'll be defining the One True Build as a farmer, almost certainly a fire farmer. Oh, sure, in theory you can balance by AT/primary/secondary... But now you've increased the amount of work required exponentially. Brutes have 216 different potential combinations of primary/secondary - add in power pools (of all flavors) and your potential number of builds to balance soars into four figures. (And you can't ignore power pools, they're a popular place to mule IOs.)
Even the Live dev team (larger than the current one and likely larger than any possible current one) couldn't keep up with keeping even one AT completely balanced within itself... let alone against all the others. That's why, after multiple Issues across a span of years, we're even having this discussion.... because, as the game is currently played, tanks are disadvantaged in group play.
And that's not even mentioning the question of what level you start balancing around IO's... Twinking is a thing, and I'm seeing heavily IOed builds as far down as the low teens.