Thanks for the response. You've got it exactly right: the sequence in which I've done the sections follows the sequence in which ideas have occurred to me. The danger of this approach is that you may be left with a residue of sections for which you don't have any ideas at all... but I think I'm okay.
You raise a very interesting question, about what kind of thing hyperliterature suits best. I don't think there's a simple answer to it, because the form is still evolving. My own preference is to avoid great big projects which would take months or years of work before there was anything to show. It's all right to do big projects, but you have to do them in small chunks, so that they can be released bit-by-bit. What this amounts to is a form of serialisation, and it has several advantages: it's less deterring for people to be drip-fed in this way than just presented with a big indigestible chunk of new work all at once; it allows the audience for the project to build up as the instalments come out; and I find, on a purely personal level, that it suits my way of working, it's got a nice rhythm to it. But just because I favour this way of working doesn't mean it's the only thing hyperliterature can do - other people work in other ways.
I hope you enjoy Stevens' poetry. It's good stuff, but it can be quite difficult to grasp at times.