The Oxy example is used to demonstrate that stats don't tell the whole story.
But for the sake of clarity, lets just focus on stats:
Having (slightly) higher stats than someone at a point in time doesnt mean you are a better player than that person. This is because past events have influence on our beliefs, which are then updated over time with more data.
As a toy example, lets say the team that just won the NBA championship starts the next season with a record of 1-2, while a bad team starts 2-1 playing vs same teams. Would you say that team 2 is better just off of that? Obviously not, but if the NBA champion continued to have a bad season and the other team continued to have a good season, somepoint therein you would start to believe it.
Going back to val lets make the unrealistic assumption that the Rating stat is a perfect proxy for your true skill level/impact/etc, treating it as a parameter to be estimated.
Alfajer (and Leo) have been better than Chronicle for basically their entire time on Fnatic, and Chronicle having a (marginally) better rating this stage is obviously not strong enough evidence to put Chronicle above Alfa. If this continues in the future, he may begin to have a case, but having 0.04 better rating is not enough data to update our beliefs drastically. This is how rational people think, and it's actually how we are currently trying to teach machines to think in the field of machine learning. If you want further resources on bayesian statistics and the updating of beliefs with data, I will be happy to point you to some resources.
TLDR: Having a slightly better season that someone doesn't automatically make you better than them. Nobody reasonable thinks that way