I find that it works pretty well for images which you need to convert to an Excel file (with your proviso that you check the data afterwards). Excel helpfully highlights cells which is unsure of, so if you check those it will ensure a better result.
However, for a PDF which has been produced electronically and that has the data embedded in it, Excel often makes something of a pig’s ear of converting it. For these situations your suggestion of Tabula for converting PDF tables is a much better route in my experience.