While I completely disagree with your characterization of chess as a "solved" game, whether it's solved or not also doesn't really matter. Aimbots can crush CSGO pros, and OpenAI beat TI winners OG in 2019, but at the end of the day the strength of computers is irrelevant because we want to watch sports being played between humans, the strength of algorithms or computers (mostly) doesn't play into our enjoyment so long as the game remains complex for the human players and spectators.
As for meta changes and new patches, chess kinda sees this in the rise and fall of certain openings. For example, the popularity of the London System in the recent decade and the development of new theory within popular openings (I recall the Tal Variation of the Advanced Caro-Kann being popular in a recent tournament even though the Short variation was seen as the most critical way to continue for white for a long time). And this is just at the highest level, anyone in online blitz has had to learn to refute the Stafford Gambit because it "entered the meta" after Eric Rosen popularized it a few months ago.