The best way to preserve the [more] endangered datasets is to use the RSS feeds. They have them split-out for various focuses like scarcity, source, threat level, etc. I went with the "all" feed and filtered in my download client.