I am not certain what you mean by a new web inferface. That sounds dangerously vague. An in-depth study of how well different sites do with screen readers, and how accessible the sites actually are for screen readers would be soooooo much more helpful. If you could identify the main pitfalls of Web devs, and address those, you could work on actually making content accessible. Reinventing the wheel and trying to make Web developers design with accessibility in mind for a whole new platform, and throwing out any and all work they have already done around screen readers, seems counterproductive.

First of all, if you're wanting to learn how to use a screen reader, I would recommend NVDA. It's light weight, free, and a huge competitor against JAWS. I really like NVDA, but screen readers do have their limitations. However, I feel that many of those limitations are because of sighted devs. What I mean is, a lot of developers, weather it be for websites, programs, etc, don't use proper methods to make screen readers accessible with their projects. In turn, it limits what the screen reader can do if it's at the mercy of whoever made the program it's trying to read.

> In an ideal world, my adviser and I would create a prototype of a web interface that could function as an alternative to screen readers.

An alternative in what way? What would be the substantially different functions? What would be the benefit?

My primary question here is "why do you think there needs to be an "alternative" to screen readers if you don't know how screen readers function?"