The concept of a DB excerpt for testing purposes is great. To make it as useful as possible, it should probably be a dump of a smaller number of feeds to one "feeds" excerpt DB, and then all the podcasts from those feeds in a second "podcasts" dump. So then any would-be tester can end up with the ability to build a complete, coherent subset pair of tables from which they could test offline, experiment, etc.
The concept of a DB excerpt for testing purposes is great. To make it as useful as possible, it should probably be a dump of a smaller number of feeds to one "feeds" excerpt DB, and then all the podcasts from those feeds in a second "podcasts" dump. So then any would-be tester can end up with the ability to build a complete, coherent subset pair of tables from which they could test offline, experiment, etc.