What they're doing is exactly what they should be doing (however - they should be doing MORE, will get to that later):
-
They are addressing Overpowered stuff in ANY group first
This is the right mindset. Master Yi can't see play in high elo without being completely busted in low elo, so it makes sense that they would balance him around low elo. Ryze could never be strong in low elo without being completely busted in pro play, so it makes sense that they would balance him around pro play.
Put simply: The negative effects of a champion being extremely overpowered are FAR worse than the negative effects of a champion being extremely underpowered. You SHOULD be able to depend on people not picking a champion that is extremely underpowered (and people that do will quickly drop in ranking until their poor pick is no longer a hindrance to their team). You CAN'T depend on people not picking a champion that is overpowered - in fact it's the opposite, you can count on them picking it. Further, you can't require players to use their bans on these things, as it means you can never ban counters or things you simply dislike.
-
They're addressing stuff that's underpowered in ALL groups second.
Basically, every champion should be balanced SOMEWHERE. It's a shame if you really love Master Yi but are in challenger - yes of course. However Master Yi is still worth having as a champion if he can be balanced in average play.
-
They're attempting to bring champions into viability in multiple groups last.
If everything in 1 and 2 is done, then this makes sense. It's a risky area (you might accidentally break the champion in the category it's currently viable in), but it's the only direction to continue to improve at that point.
Now - as to what they should be doing more:
The big problem is with their criteria for balance.
It's not simply enough to identify a champion as overpowered in average play and top-10% due to extremely high win rate, or a high win rate with a high ban rate.
I've consistently argued in the past that a champion's playstyle, place in the meta, and difficulty to play, should all factor in to balance. The article implies these things don't matter. Some clear examples of this:
- in almost every meta ever, any AP damage comes from mid lane. AD Mid laners should satisfy a very useful role when they have allied AP damage, and have every right to be strong niche picks in these games. However AD Mid laners (especially Yasuo and Zed) are often so extremely popular that they are picked regardless of team comp. If this is the case alongside a >50% win rate, they are Overpowered. In order for them to have an over 50% win rate while being picked so often, means they must have an over 50% win rate in suboptimal games (e.g. 90% of their games are on all AD teams where they SHOULD be at a disadvantage). FURTHER, this means they likely have an EXTREMELY high win rate when they are in an appropriate team comp (making this up, but e.g. they might have a 60-70% win rate when the allied top lane is an AP damage dealer like Sylas or Vlad). This makes them super abusable in those cases, but more importantly unfair to people who have to play against them. It's NOT FUN to lose lane to a champion that's overpowered even if you have a 50% chance at winning the game because your tank can stack armor and become unkillable due to them not having magic damage.
- a champion that is difficult to play shouldn't have a >50% win rate in low elo alongside a high play rate, especially a high inexperienced play rate. This implies that the champion is balanced around inexperienced players, but being difficult to play will be much stronger than this with experienced players. Players who take advantage of this state of balance can easily climb far above their decision making abilities due strictly to the power of the champion and their ability to play it well mechanically. When the champion is eventually moved out of favor, these players will be far weaker than their current elo and this issue is likely the source of people that you question why they're in your games.