Our Web Service is not a traditional API; it is intended to be the back end to a website. As such, it is not intended for downloading all of your data. (You can download from it/cache some data, but this should be limited.)
The Web Service can return up to 25 results for a generic query. When you send a request for detailed data, you will receive one title result for each hit to the server. Since the Web Service is not a continuous feed, you would need to check back periodically to receive updates to the information if you choose to cache data.
Note: We do monitor for large groups of consecutive hits that appear to be created by an automated process. We ask that there be at least a second between your requests to our server.