Looks like you have to reference it in the comment, rather than the summary. So, this is for #5267.

Alternative solution: instead of replacing the PUA symbols with other Unicode symbols, we could replace them with the ASCII characters you type to get them. E.g. =>, <>, etc.
Advantage is that they are more likely to appear in the user's braille table.

This looks like a good resource for mapping from the Wingdings font Word uses for some bullets to standard Unicode. Remove the 0xF0 prefix from the PUA codes and look them up in the table. Unfortunately, this doesn't quite match what is done in the PR. I will discuss it also with sighted users.

In Word you can also Save As --> Plain text --> UTF-8. This way Word transforms PUA codes into Unicode, although the result is not unambiguous. E.g. U+F0B7 and U+F0A8 are both transformed into U+2022 (bullet). Still, in my opinion it is a good idea to do what Word itself does.

The only problem is that you can no longer identify each individual list level, because some will share the same bullet symbol. But this is already a problem if you make a list with more than 3 levels, because then the bullets start repeating.

Closing this because #6778 supercedes it. This provided a valuable basis for the other PR, but I don't think we will be merging this one as-is. However, please reopen if I am overlooking something.