It can be argued that most pieces of electronic literature could be reproduced in paper, thus the question about essential innovation seems valid. What cannot be reproduced on paper is the processing capacity of a computer. Storyspace offers some basic processing. Literatronic offers a sophisticated IA engine for processing.
I’d like to clarify one issue. In my view, Storyspace, to pursue the path of aesthetics, allows for the writing of hypertexts appropriate to its environment. In all honesty, I don’t think enough hypertexts have been written in the software to provide fuller analysis of the possibilities: I’d suggest a thousand (how many books are published every year?) In this sense, appropriateness (a criteria I take from urban design), is a key factor. This is a subtle but important point.
In Brimmer and Death, for example, I’m not worried about links or the relationship between one window and the next. Rather, I’m fussing over regions of the writing space: top and bottom, primarily, for reasons of closure. This will put into motion a new editing stage: given the top of a writing space, how should I fuss over the bottom of its origin. I’ll be looking at this very issue as I experience a show of hands.