Throughout history, players have been on the swing of following the “mainstream” dogma, with some brief periods of breaking the mold and revival from time to time. For thousands of years, fixed stone opening is the norm, or even the rule, where different regions have different number of them and how to place them, and we eventually break from them in the Edo period Japan.
And then after hundreds of years, the starting with komoku (3-4) and a series of fuseki where players who don’t play 3-4 opening would be unthinkable and games from that period showed they repeated several identical opening fuseki based on the idea, and continue till the 19th century.
At the end of 19th century, players start to experiment more as the traditional houses no long have the power to restrict how they rank players, and finally after decades we have shin fuseki flourished in the early 20th century, and then we keep experiment till the 80s mostly and different styles or schools in different regions start to settle again, and reach a pretty high level in the early 21st century, where some openings were considered obsolete and if not for AlphaGo, the pros now might repeat the pattern like they did in the past, where Asia pros stagnated and what they considered good fuseki will be the mainstream now, and players across the world likely start to copy them (since without AlphaGo, the popularity of Go might not spread that far, and resources will still be limited to East Asia sources and pros)
Luckily, we have Go AIs that add more possible openings and even give evaluation value to them, but still not good at judging human capabilities and our grasping the openings (some opening with even 20/80 split, still playable even solid in human pros’ eyes and almost certainly balanced for the rest of amateur players). So in a sense the breaking the mold happened at the time it was supposed to solidify might be the cause of players to remember the good old days. But in reality if an opening is good and useful (even the cross fixed stone placement that can prevent mirror go strategy are useful), and old solid opening fuseki still used and balanced in AI’s perspective as well (like no different that tens of a percent different). Just as pros are still “exploring” these AI openings more in their games, doesn’t mean the solid and good ones which stand the test of time are not there. They accumulated and stay in the collective knowledge. Only that pros like to start with fuseki which they think the opponent might not know some variations or positions following them and will be in their favor, so they pick these new AI fuseki as their tactics, and they will start to explore more and more, it is in their nature (as many younger pros start to show their different tendency in openings and for the most part workable and balanced new openings)
I would expect that when the generation who started with AI dogma into the mainstream, there will be truly new explorations for even more variations when they already have sufficient deposits of all kinds of fuskeis in our collective knowledge base.