[AriaNotify] Speech markup/pronunciations (i.e. SSML or IPA)

[alisonmaher]

As discussed in https://github.com/w3c/aria/discussions/1958#discussioncomment-6223153, there may be cases where an author would want to provide a separate string to provide a hint to ATs for how the notification should be spoken.

The use cases provided in the discussion thread are compelling, and I think there is value to such an addition. However, I would like to propose that we consider this for a future version of the API instead to allow us to gather feedback from developers and end users as ariaNotify starts to get adopted with its current set of functionality. This will help inform the best way to incorporate this additional customizability going forward.