I got really excited about aural style sheets back in the day, only to have my hopes dashed when I discovered no one supported them. Then came ReadSpeaker (and its ilk), which started to move in that direction, but the solutions are far from perfect.
I even tinkered a bit with the Speech Synthesis API to achieve something akin to a functional listening experience for this blog. Would love to go back to doing it in CSS though. I hope this happens!