I love itch and have enjoyed a lot of games I've bought on here. I've found some interesting gems both on the side of artistic trashgames, and of commercial fun machines. However, I've frequently found these two particular aspirations clashing with each other in search results, making it much harder to find specifically the type of game I'm looking for. This has by far been the biggest thing preventing me from exploring itch more, and the reason most of my discoveries have been through twitter and mastodon, because these two goals are almost exclusively in opposition to each other, at least at this time.
I have two ideas that I believe would help untangle this and give both aspirations room to breathe.
One option is more heavy-handed but as a result, probably much more effective: Straight-up introduce "Commercial" and "Artistic" categories, and allow games to place themselves in whichever one they are aiming for. These wouldn't be mandatory or mutually exclusive, there can still be a general "browse" page for all games, and search can still search both categories, but there would be prominently displayed and easily accessible ways to limit browsing and searching to that particular category, in a way that stands out from tags and genres. The primary issue I can see arising is that this is fairly limiting, and in future, these categories may want to be split into more, or there may be more added, or perhaps they'll be torn down altogether, but I think at least right now, it would make the experience of finding games much more reasonable.
Another option is a much softer version of this: Official, canonical tags endorsed by itch, and presented prominently when making a game's page. The effect would be similar, but less heavy-handed. There wouldn't have to be a discovery process and a lot of networking to figure out what lingo people are using for their tags to delineate what type of game it is, just easily visible tags that can be displayed in the store to customers and creators alike. This has the advantages that it uses an existing system, and can easily be expanded to include other things in the future.
Thoughts on this? I feel pretty confident in the need for some kind of solution, and I don't believe curation nor black box algorithms are good ones.