I've dug into the code to figure out why this might be happening and have narrowed it down to the function _importexportapi_db_put_generate_key_ids().

It appears that the content type name is coming from the <name> element inside of the <story> element of the node_data.xml file. The <name> element is supposed to be the authors name of the story, but the importexportapi is pulling that as the name of the node_type entity.

I've looked through the definition files importexportapi_node.inc and importexportapi_story.inc and there is no "name" field in the definition for node entities, but there is a "name" field in the definition of the "node_access" entity. For some reason the node entity definition is pulling definitions from the node_access entity. When I'm running the import of the test data I am not choosing to import node_access so it seems that the node_access entity definition should not even be loaded.

The _importexportapi_db_put_generate_key_ids() maybe needs to be re-factored into smaller chunks or better documented because something weird is happening in there in regards to this. From what I understand of the function it takes the fields in the definition file that are marked as keys or alternate keys and makes sure they exist in the database before the rest of the entity is imported so that the entity can reference those fields.