Acrobat accessibility character encoding fail on certain languages, eg Punjabi, Thai, Khmer

At the agency where I work, we create a lot of PDFs from Indesign that are in other languages. When I use the accessibility checker on these in Acrobat, I've been having a character encoding fail on several languages. They seem to all be ones where the characters include little flourishes either below or above the character – so like Punjabi, or Thai, or Khmer. Typically we use Noto Sans from Google, so it’s definitely unicode, and of course we are using the world-ready composer, and exporting with the correct language assigned. All that required stuff. It’s not solving this.

Does anyone have any suggestions?