The problem is not with the metadata download plugins, it is that you undoubtedly attempted to download metadata in bulk and did not have a sufficiently strong starting point for those books like having the ISBN (often extractable using the Extract ISBN plugin) or non exact title/author.
Without the ISBN, any metadata download plugin is at the mercy of trying to parse the search results from the website, which depending on the quality of their search engine and how they order the results can range from "not a problem" to "not going to match".
In terms of bulk deleting ids, I can't see the point personally because it is only going to tempt you to bulk download again which is what got you the problem in the first place. To correct it, you should be one by one clicking the edit metadata download button, click the "Clear" button to the right of the ids field and then while in the dialog click the Download metadata button and *manually* choose the download result which matches the book as you know it. Sometimes you need to click the download metadata button a second time to get all the ids you want populated for the same "book" from the other sites, because after the first download it now will have an ISBN set that the other sites can match against.
Bulk metadata downloading "can" work, but it is preferable if it has an ISBN to work with. And even if I do a bulk download I *always* double-check the results to make sure it is linked to the right book and revise it using the single metadata download as discussed above.
As for your "defaults" comment - Overdrive is woefully slow, which is why it is not a default. I don't personally use Google. None of the others you mention ship with Calibre (I wrote them and use all of those). As for B&N - it provides better quality covers usually than any of the others (except for BAEN books via the Webscription plugin).
|