The commitments and expectations required in hypertext literature now assume a level of technological accessibility that does not exist. Maybe, in the next several decades, computers will be as ubiquitous as books are now. Maybe electronic reading devices that can use the links, structure and programming electronic literature is exploiting will be mass-marketed. Maybe the price will come down or the libraries will open up or we will have charities that understand the gaping need for electronic media. Maybe we will be smart enough to fund more efforts like the hole in the wall (Judge 2000) experiment which gives slum children in India a computer to access. Maybe. But don't hold your breath.
Moreover, by pushing the edges of technology, hypertext writers keep limiting the potential audience. We not only require literacy, but access to an ever changing array of machinery and software. Writers explore the latest tools, requiring readers to have ever expanding hard drives for Flash and Shockwave plugins, sound systems, screen bandwidth. Writers lament the loss of works when Netscape 6.0 does not support DHTML, VRML virtually disappears, HyperCard players no longer come standard or supported, floppies become a thing of the past, and InterMedia fades into an inaccessible legend.
You can go to any library and get a copy of Dante's work 700 years later. You can even get a copy from Project Gutenberg that will run on any computer, Palm Pilot or electronic device. Who can access hypertexts now? Who will be able to access hypertexts just 50 years from now? And how do our current access and individual technological quirks drive our reading? Is archiving an ontological criterion?