Well, what else do you expect from crafting? How would you design it to avoid some sort of skilling up? You know what takes a lot of grinding? Art. Actual blacksmithing. Computer programming. Music. Anything crafty or artsy is grindy as shit irl.
How are you going to simulate crafting in games? Crafting is something that very few people do in real life because its hard to do and not fun. You have to have that drive to push through the massive tedium while making nothing but shit for years. I think you love the IDEA of crafting, to paraphrase my japanese teacher, but you don't actually enjoy creativity.
Do you imagine that's its possible to strip out the tedium of experimentation?
Perhaps you could explain what you thought crafting would be like, so it would be possible to figure out if what you imagined could be replicated or if its just an impossible fantasy. Not mechanics, but the abstract experience of crafting that you see in your mind.
The thing is, the tedium is very much NOT in the experimentation. To the contrary, the tedium is in gamey crafting being extremely... linear? and being so far removed from creativity, innovation and uniqueness that it might very well approach it from the opposite side. Like others mentioned, take Minecraft - the actual creativity begins ONCE you already have shit crafted and start doing Stuff with them.
You brought up music - funny, I'm actually pretty seriously into learning not only my instrument but also how music itself works - and guess what, the only actual tedium? Physical exercises, and from my experiences you can learn all the skills naturally by playing stuff that is only slightly above your level (as opposed to being way above it, or on par, because then you don't learn shit).
Art is a little different than craft, although I stand by my claim. You improve in craft by making the same shit over and over until you do it right, until your muscle memory can feel that its done right before you finish. This is completely impossible in video games. They can't model physical laws like that. Well, they can but that costs serious money to make a game that way.
People complain a lot about RPGs, which most MMOs are, being all about the character's skill and not yours. Like no shit guys, do you not understand the definition of RPG? Role playing, pretending to be something YOU aren't doing things YOU can't.
There are a lot of things a game would need to be like real life in crafting. First you would require some sort of physical interaction. Even if you have a game with hidden recipes people set up stupid fudging wikis and spill all the deets, ruining the exploration aspect. This is the curse of knowledge. Its too easy to pass around digitally. So you need to force people to do something complex in the moment. Something dynamic. And then you hit the reason roleplay works like it does. Most people will never under any circumstances be able to do top tier stuff. No matter how hard they try. Success will depend on your inborn and enviromentally given traits and development. The whole point of abstract roleplay is that your physical real life limitations don't apply. Be they mental or body based limitations.
This ties into the story of video games as a whole. The video games of years past never broke into the mainstream for the most part. The video games the majority plays are nothing like the games I played or imagined as a child. Similarly, a video game that involves complex mental tasks like involving math or reading or memorization will never be popular. I would be incredible at such a game. But I would be the prince of nothing. I was a master of Tactics Arena Online and of Warring Factions but no one gave a shit. Unlike the approbation of friends when you rule at games that plebs actually enjoy like shooters. And its not about my vanity. Its about money. Games involving mentally complex tasks lack accessibility and thus a user base to make it worth developing them.
MMOs require serious bank to be made.
Aside from the difficulty of making a game that couldn't be fucked over by a wiki detailing how to win at life, we have the trouble of content. Hand crafted crafting of substantial depth would take shit tons of time and money. And procedural crafting would be incredibly unbalanced in a multiplayer game. And further it would require an entire sandbox world, you can't make story based games when the designer has no way to match your power level to the difficulty level of the content.
Look at even mildly sandbox games like EU4. The devs are constantly shitting on the sandbox because a few people, including me, can exploit the crap out of it. I've caused rule rewrites for entire text based games and I'm not even on the level of dedication that a DDRJake brings to the game of bending reality, or virtual reality/fantasy I guess, to his whim.
I spent a long time developing an MMO concept that could throw down against wikis and control min-maxing to an acceptable way, and for clarity, you can still min-max like a beast. I got some pretty good responses to it, too, even though it was explicitly just mental masturbation and not an actual product in development. And even though I spent years before and after the year I really got into it on my ideas for ideal MMOs I couldn't dodge the grind. Aside from perhaps my potion making system.
You would fight a dragon at high cost to get a sample and do experiments, combined with your knowledge of herbs, and magic, because magic was intimately woven into crafting in a way no current game does, to derive more and more effective poisons to make bringing down dragons much easier. For instance for one dragon you had to figure out a metal particle that would bond to their blood, sorta like how hemoglobin carries oxygen, and connect it with an anti magic material to break down their natural poison resistance and then apply poisons that could deal with their specific biology. It implemented a slightly simplified version of real life organic chemistry to create and describe reactions. And even with that, my crafting feature masterpiece, I still only managed to solve 70% of the issues I have with crafting in games, much less solving the issues other people had that didn't bug me.
I even had some working, but text based, systems written in C++. I couldn't imagine writing an actual GUI that could handle the dynamism of that emergent system though. Did I mention it generated the rules dynamically? And without a database or debugging tools of much note I actually didn't know myself what would kill the example dragon I caused to come into, again text based, existence.
And I'm sure with time some jerk who is even smarter than me could develop a pattern to predict things, give it out on a wiki, and cause my elegant magnum opus to crumble into dull tedium yet again, defeating the entire goal of the system. And in every game some altruistic shitlord will figure out and wiki your stuff for the masses.