Results indicated that the accessibly designed site had equivalent usability in the graphical browser, but was significantly more usable in text-only and small format browsers. The results support accessible design as universal design, of benefit to all users, independent of disability.
I've always thought the single most important thing is to MINIMIZE THE NEED FOR MOUSE USE.
This relates not only to blind people and those with arthritis and so on, but also to the able-bodied who are trying to minimize repetitive strain injury, or keyboard-oriented people like me who just don't like using the mouse more than is absolutely necessary.
Thus I have always tried to design my site so that it can be navigated easily without a mouse if necessary.
This is among several reasons why I never use frames or drop-down menus or java scripts, among other things.
It is a great pity that so many people, when they learn to use a computer, are never told that most of the things you can do with a mouse can also be done on the keyboard. Training courses for software programs tend, in my experience, to perpetuate this problem because the trainers only ever tell you the mouse way of doing things.
The article is a good counter-balance to the argument that an accessible site reduces the options for the webmaster and removes features and functionality.
There is still a lot of resistance to concentrating on accessibility as the benefits seems vague to many, so the fact that improving accessibility will naturally have a positive effect on general usability makes the effort much more worthwhile. So we can move from "I don't care about accessibility, I care about my users" to "I care about my users so I care about accessibility".