A little bit of background (feel free to skip):
There are a lot of government regulations being put in place to prevent issues with website accessibility under various anti-discrimination acts in the US, Australia, UK and other countries around the world. To put laws in place to prevent discrimination creates a type cost for companies if people sue them for not being able to access the information easily, and that has been the reason why some government agencies in certain countries are making an effort to comply with WCAG. However, this type of incentive leads to websites that aim to comply with the standard, and not delivering the best user experience possible.
Having multiple sensory input available makes it much easier or faster for us to perceive or interact with both the digital and physical world, as there are either redundancies or synergies with being able to perceive more than one sensory cue. Take away the ability to perceive by sight, sound, touch, etc. and you have to develop more acuity in that sensory perception to compensate for the loss of one or more of the others.
Looking at how accessibility is provided in the physical world for a moment, and see how it compares to the digital world. Take a university campus for example, there are access to different buildings in the campus, access into buildings to consider, and access within the confines of the building as well. Access to different parts of the campus and different areas within the same building are different accessibility problems to access into a building, and so different strategies are employed. Creating a disabled parking spot only reduces the distance to the building, not how easy it is to drive around as a disabled person. The point here is that in the physical world the problem has to be about making a car drivable by a disabled person and not creating more disabled parking, but it doesn't impact on the rest of the campus layout so the impact is less significant.
Entrances to buildings are not a hybrid form of a ramp and staircases, because generally to make ramps accessible for wheelchairs they have to have a low degree of inclination whereas stairs are designed to create a quick way to bridge vertical distances. When there isn't enough physical space to build a ramp with the correct slope then it requires a mechanical device to move the wheelchair. The point here is that in the physical world this does have an impact on how the entrance to the building is designed, and the cost of trying to build it into an existing infrastructure is always more costly.
Some of my thoughts (again feel free to skip):
When weighing up the cost (not just the cost to the company, but also the cost to the user) and benefit of designing an inclusive website that caters for all types of users, is it not better to separately design and build a website that will cater for users with normal access issues? Or should the improvement be made in the screen reader or devices to try and support these types of users? I would definitely love to get an opinion from someone who is a vision or motor-skill impaired designer/user (I wonder how accessible the UX SE site is as well), but short of that I like to see if there are those out there that have extensive experiences with screen readers and accessibility requirements.
My personal theory is that most of the modern day websites have been either too complex (e.g. rich, multi-media), the information architecture almost non-existent or not thought out well, the user of flat-design introduces too many subtle interactions and that we haven't spent nearly enough time or effort investing in technology for vision and motor impaired users.
I am interested in any thoughts from people who had to deal with accessibility issues for compliance reasons, and whether they were able to justify the cost of creating a separate website on the basis of taking too much away from the usability and user experience of an existing website by trying to make it accessible to such users. Perhaps this is why only government websites are creating a mandate for accessibility compliance but the commercial organisations have not made this the same priority.
So the questions is:
Would creating an inclusive website actually take away the full experience from 'normal' users while not creating a website specifically based on the needs of 'disabled' users also reduce their experience using the website? If so then does having just the one website outweigh the benefit of creating a better experience for all users?
Is this the same issue with responsive design seen as a solution to creating a consistent experience, although many companies building mobile-specific websites because they have mobile users with a very different need or behaviour to desktop users?