One thing I've noticed about my own work in contradistinction to Matt's and Steve's is that I don't have a metaverse. Matt's works all tend to be "backended" into each other as part of some larger shared reality, and Steve's are (so far) in a shared universe that's meant to be expanded in multiple directions. With mine, every work is wholly its own thing, with no connection to anything else.
I call it the "Simak model", from how Clifford Simak seemed to take this approach with his work. Every time you opened one of Simak's books, you were starting completely from scratch, as was he. Contrast this with the "Heinlein model", where most everything he produced could be located somewhere within some extended timeline.
From time to time I've flirted with the idea once upon a time of making a metaverse or extended Heinleinian storyverse, but I ran into a few problems every time I tried to map it out.
One, I could never come up with something open-ended enough to satisfy me. I was always conscious of how it might be harder to fit later work (from a more advanced stage in my development) into the framework developed to support earlier work. Just having the later work happen later in the chronology might not fit at all. I know that we all make it up as we go along, but I didn't want to end up being hidebound by earlier decisions that were made by a less enlightened me.
The other problem I ran into: I liked the idea of having no legacy strictures on whatever story I was currently working on. If I could guarantee that something of mine had no inherent connection to anything else I was doing, then I didn't need to figure out how to backend anything into the larger framework. I could just do it.
The third problem: What would it really mean to say that all these things fit into a bigger timeline, other than that they ... well, fit into a bigger timeline? It didn't really add anything to the stories themselves. The idea is, as I see it, that you can use allusions and connections between works as a way to enrich all of them at once. But I had a hard enough time making any one thing work on its own, let alone figure out how such things could be linked in a bigger picture.
Note that I wasn't trying to apply this criticism to other works created in this spirit -- more power to 'em! It was my own work that I couldn't see benefiting from it in a way that mattered to me. No matter how I sliced it, I couldn't get anything out of it that didn't just feel like a gimmick for its own sake.
I know the idea is to have inspiration and source material from one's metatexts, rather than see them as restrictions. In the end, rather than struggle against the odds of something I wasn't all that inclined to do to begin with, I decided to just make everything its own thing. It also meant less bookkeeping overall: each work could be tracked and managed separately, with no worries about how they had to be consistent with anything else.
One of the more Taoist insights I've taken away from creativity over the years is that everything that seems like a problem is actually its own solution. If you have a difficulty in a story, that might be a sign the difficulty in question is the story's real subject. I thought about this while struggling with the extended-universe concept, that by making the idea work for me I would be transcending a limitation in my own creative process, and enjoying some additional growth. Now, I'm not sure that's true. If the net result of beating my head against that particular wall has been nothing but a bruised forehead and some dented sheetrock, maybe that's just a sign I should not worry too much about this not being a thing of mine. It's not as if I don't have plenty else to wrestle with!
New York City
Other Lives Of The Mind