I use scanbot on iPhone which has OCR. After I import into DT3 it shows up as PDF+Text and I am able to search it. As an example, original file size is 1000KB.
Then I use DT3 for OCR and file size drops to 337KB.
Question is, does it “hurt” the run OCR on pdf that already has text layer? I am not sure which product has better OCR (I would guess DT3).