> Yes, you need an internationalized list of strings to detect.
Who would maintain this list and be the authority for every language on Earth?
We've managed to get this far without needing such a central dependency.
It's not a "central dependency" that needs an "authority". It's just part of building internationalized software.
Shouldn't screen readers have intelligent heuristics to most appropriately convey context when required? Seeing as most of the web doesn't have accessibility annotations?
There may not be a surrounding sentence (e.g. in a "PDF"/"Download" link). The surrounding sentence may not add any/enough additional context, e.g. "You can click here for help." or "To view the article [one of many] you can click here.".
You then run into issues where 1) the screen reader/AT is overriding the ARIA/a11y markup on the page against WAI-ARIA and WCAG standards; and 2) you can end up with different behaviour on each screen reader, in addition to the places they already differ.
It's bad enough when web browsers introduced heuristics on form labels such that "name" + "label" fields were detected as a login form and other situations where bad heuristics were used and the web browsers overrode what websites were specifying.
It would be an extraordinarily easy for screen readers to have a heuristic that whenever a link is just "click here" or common variations like "tap here", "click", etc., to read the entire sentence containing the link. It's not exactly rocket science. Yes, you need an internationalized list of strings to detect. Also, if aria-label is present, just use that.
Likewise, search engines are great at inferring meaning from the page as a whole. I'm not going to change my link text for the benefit of search engines.