The major concern with your approach, for me, is that you're very neatly compartmentalising design away from reproach or criticism. The game ends up sucking? It's the coder's fault for not achieving the vision. Game not finished on schedule? It's the publisher's fault for not hiring enough staff to generate the content that the vision demands. Reviews terrible? Well that's not the game I had in my head that ended up onscreen so don't blame me. And so on...
This is true. But only in the role of 'visionary' as a director or a lead designer, Those are all your fault. You should have pushed them harder, you should have been the one making it happen.
That assumes the idea wasn't bad to begin with. But I am a firm believer that you can make anything happen. Miyamoto is a testiment to that.
I don't see, from what you've said, how you envisage the design ever being analysed on a cost/performance basis. It just sort of sits on this holy mantel, away from mere concerns such as practicality of implementation and market viability. That might superficially sound like putting the design first as you auteur excellence, but what it really does is cordon design off to a point where it's in danger of drifting into irrelevance. I'm sure you can recall what happened the last time someone declared design as law!
On the contrary. These are all factors that go into makeing a good design. If it can't be made, or no one would ever buy it. You failed. The design is bad.
I'm not surprised that we don't have really good noob-tools for making fun games. It is romantic to suggest that the optimal designer for a game is the person who will be playing it. But no one wants to explore that in reality. Game designers enjoy being creative. They also want to impress their own values and ideals on others.
You can't learn something from a game you created yourself. At least, beyond self-reflection and technical skill. You can learn something from playing a game like... Final Fantasy 6. Or Harvest Moon. Or even obvious games like Where in the world is carmen sandiego.
There is little interest in designing a really robust snap-together gameplay system with procedural content generation and random-but-not-really-random events and boards.
The gameplay should be designed by who is playing it. That doesn't mean the content should be. You can never be surprised without that aspect, and therefore, you will always be bored playing your own game. Unless it is a game like a puzzle game, or with a lot of emergent situations.
You need something like the holodeck. It doesn't have to be holograms, it can be on a screen. It doesn't have to be voice activated. It can use a paragraph, or even snap-together details. But it has to be able to surprise the player. So it needs some kind of bank of ideas.
This is why these kinds of things don't exist. No one wants to make them, they are more advanced than current technology permits and they would be very very hard to balance.
I did have an idea tho. You could have a bank of potential ideas and mechanics. An online listing of 'random' generated games based on those, and a rating system and variation system that enables the game itself to fine-tune it's own gameplay using the players as a measure.