Personally I'm skeptical it wasn't just a product of context. The Xerox document model was logical, in a world that still heavily relied on paper. Having an abstraction that seamlessly interoperated with the century of information storage that preceded it is a no-brainer. In today's world where paper is becoming increasingly rarer, and much work has been done in digitizing that mountain of documents? I'm not so sure. And I think Engelbart was focused on that future, rather than the 30-year transition period that would end up happening.

It's not to say that the specific implementations Engelbart was working with were good. But I'd point to Plan 9 from Bell Labs as a kind of hybrid between Douglas Engelbart's vision and what Xerox produced. It's a little alien, but relatively easy to learn, and at least conceptually it shows that an unstructured UI made up of hypertext and windows can be quite nice to use. When that's integrated with the primary IPC mechanism of the operating system, which also happens to be the filesystem, you end up with an intense synergy that's hard not to be delighted by. I don't think it was possible to avoid computers becoming digital filing cabinets, but I also don't think we should write off moving beyond this era at some point. There is a large, underexplored dark wood. I am very interested in what lives inside. I think revisiting Engelbart's ideas of human augmentation with a prolog-based system like the Japanese Fifth Gen Computer project has extremely promising implications.