Balancing Strategy - What would you prefer?

Balance issues seem to be a major concern these days looking at the top threads and I also agree to that. I don't think it's as bad as people are saying but there are some glaring issues that you would think Riot could solve easily. Since Riot likes us to use constructive criticism and give suggestions I was wondering what balancing strategy people would like Riot to use. There are two points to consider for this: 1- Champions will never be 100% balanced 2- People have different opinion and play at different level

We'll never get to a point where all champions win 50% of the time (Common folks) and we get 100 different champions each week in the LCS (pro players) because they have to train around something and if Urgot becomes good tomorrow after a big rework, it will take a few months before they bring it to the level of their other champions they played a thousand times.

With that said, how would you like Riot to balance the game?

1. Empirical: Riot buffs the champions with the lowest win rate and least pick professional pick slightly until they are competitive. This strategy has the advantage that at some point champion numbers overshadow any weakness that they have. Let's take Viktor in example, I don't think he ever been picked in the LCS and his win rate is around 48% right now. Riot could buff his numbers until he is viable. At some point, if he can destroy people that gets to him or if he can poke the enemy team enough he would end up being picked. Using this strategy, each patch notes would have changes to almost every champions by buffing them slightly or nerfing them slightly.

2. Roller coaster: Riot completely nerfs the champions that are always picked in the LCS or have insane winning rate and buffs significantly champions that are not picked at all. This would cause major changes in the meta each patch. This has the advantage that we would see each champions played professionally over time and that we get to play each champions as they get their prime time. This has the problem that the game would be harder to pick up since you would have to adjust constantly. Using this strategy each patch notes would contain significant changes to the top and bottom champions but anything in the middle wouldn't be touched.

3. Reverting nerfs to champions when they get out of meta . A lot of champions used to be played competitively and suddenly disappear after they are nerfed. Here's a few of those changes: Ryze ability range, Pantheon and Nocturne ultimate distance, Zac tenacity during his ultimate. This has the advantage that there is an easy proven way of making those champions competitive again, or at least they can be tested against the current meta picks. This has the problem that it wouldn't solve champions that are not competitive because of their kit.

4. Current method which could be described as slowly moving. This has the advantage of having similar picks viable for a long period of time so people can learn champions and perform well on them. This has the obvious disadvantage of having dozens of picks that are completely unviable.