_Note: I initially posted (a version of) this in the comments below the framework before realizng that it was unlikely to ever get read there. I'm usually opposed to this sort of "visibility motivated cross-posting" that doesn't actually broaden the platforms the content is available on, but it felt appropriate in this case. Mods are welcome to correct me if they disagree_
I love the attempt to provide metrics and use hard, data-driven justifications for balance changes. I think that there's a bit of a risk of the larger community lacking the knowledge and experience to understand that such an attempt is always incomplete and flawed, as there is no way to make a literal translation from raw data to specific actions without introducing _some_ level of bias or making _some_ sort of interpretation, but the cost of that is some rioters getting yelled at by the more impulsive and less thoughtful members of our illustrious community, and the only way that would be new is there might be a slight shift in which Rioters in particular get yelled at.
however, I would argue that trying for meta-agnostic balance on a champion-by-champion basis is the wrong approach, and one that has caused balance issues before.
I would expect that making use of current champion archetypes (e.g. "Marksman," "Juggernaut"), adding a different set of categories that describes specific play-patterns ("Tank-buster" is the obvious example, but I could see taking "Sweeper" from Pokemon (a champion that if set up with sufficient gold and a proper engage has the potential to wipe or nearly wipe the enemy team, such as Katarina or Shyvanna, these champions excel at punishing teams with a poor frontline)), "Split Pusher") and using and data on the meta to predict how a champion "should" be doing, then balance based on how far away they are from that mark.
Otherwise you end up applying buffs and nerfs that will push a champion away from balance the second the meta shifts, as well as broadening your targets to every champion affected by the meta (read: all of them), rather than a more achievable target of those champions over performing or under performing in the context of the meta.
With the current philosophy, you would be expected to buff Vayne and Sejuani to compete in the early game damage meta while nerfing Lucian and Lee-Sin, then turn around and do the opposite when things slow down. You'll be constantly chasing your tail; this will not only prevent champions from slotting into their unique strengths and weaknesses, but will also reduce pre-game strategy, artificially shorten metas, weaken the player base's ability to adapt to metas, and worst of all, the focus on champions highlighted by the meta will prevent the slow improvements in systemic balance achieved by slowly improving all the most consistent offenders.
Mind you it's not an accident that I mention Sejuani and Lee Sin, two of the more notorious examples of the risks and limitations of meta-agnostic balancing.
I understand the temptation, it's surely frustrating to be yelled at about Lee Sin every time there's an early-jungle meta, but I think this approach represents scope-creep and will sabotage any effort to make real, lasting balance improvements.
**TL;DR** (I don't blame you): Using hard metrics that ignore current metas to determine balancing decisions might be the simplest way to justify them, but it has many unintended side-effects, the most obvious being that champions with an intended power niche will be trapped in a constant cycle of meta-determined buffs and nerfs, preventing Riot from spending quality time on champions that are consistently somewhat over- or under-tuned, and limiting Riot's ability to make lasting balance improvements.