This has become a highly valuable thread. I've learned a lot, so thanks!
Karen Coyle wrote:
Having bib records without any way to get to the item isn't very useful. I can't remember the number of holdings that OCLC cites, but clearly it is the holdings information that sends people to WorldCat. And, of course, the essence of ILL services is knowing what libraries have an item.
One of the most important customers of WorldCat are catalogers and order staff, where being able to get a copy record into the workflow as soon as possible saves a heck of a lot of time, and therefore money. Perhaps the item hasn't even been published yet, but there still needs to be a record for it. So, having the bib record, even without the item, can be highly valuable.
Of course, much of this is based on standards, because if a record is so badly done, or is following completely different standards than what you are using, maybe everything will have to be redone and the original record is useless. The more a record has to be reworked, the less value it has. This is why I think standards and quality take on even more importance in this new world, so that a publisher could upload into a system their metadata and provide truly valuable information that everyone can accept without each library (or even a single library) redoing everything over and over again. Publishers should not be expected to provide an entire, finished record, but certain parts of it would be more of less guaranteed, e.g. certain parts of the description or specific ISBD areas. Different entities could add their parts and the record could gain through some form of accretion. This same process could slowly be extended to other projects as well, such as open archives.
But if everyone is engaged in a furious race of fixing everyone else's mistakes, we will be like a dog chasing its own tail. I admit that OCLC could be a very important part of this, but the record use policy does not seem to be heading in these kinds of directions.