Together with default multicore, it should be possible to optionally index Content and Locations in a single core Solr setup.

Document should still be indexed per language, that will require for document IDs to incorporate language information. eZ Publish entity IDs (Content and Location) should be indexed in a separate field.