There was a beta 1 with the major features present in some form except for any real UI integration and server side support.

I may be retroactively misremembering what people were saying one or two years ago, but I'm fairly sure the lack of a useful UI was one of the big problems. Just like the guy said in the post at the top of this thread, there's no way Grandma was ever going to learn how to write natural language queries. If you can't figure out how users will do that, you're not solving the problem.

So the problem you want to solve is people can't find things efficiently on increasingly-large hard drives. You come up with a solution that involves changing how things are stored and retrieved. Then you don't come up with a usable interface for people to interact with this new file system. Did anyone on that project have the end user's interests in mind?

Sounds like you're saying they wanted a reason to write a new file-system that made SQL Server more efficient. But doing that would piss off the rest of the software industry, and would (correctly) be seen as leveraging their OS monopoly to extend an advantage into the database sector, in which they have real competition. So they instead put forth the explanation that they were trying to help people keep track of content on their huge new disks, knowing from the start that there was no way it could ever work for that.

Was that what you were getting at, or did I just make up my own conspiracy theory?