Allow automatic OCR setting to be configured to only apply to file types where it is beneficial
Currently, enabling automatic OCR applies to all files. It would be great to have an option to have automatic OCR only apply to files that would otherwise trigger the 'Run OCR' option during file import. This would ensure we don't run OCR on files where it's not recommended, but also users wouldn't have to do the extra click for running OCR because we recognized a file type where OCR is recommended.

Automatic OCR will only apply to documents with no text layer. If you import a document which contains text, this will retain that text layer (even if automatically apply OCR is set). As a user you would need to manually choose to apply OCR in that instance.