Optimize entity retrieval from query result

[ledsoft]

Currently, when a query results in entities being retrieved from the repository, it leads to (at minimum) n + 1 repository calls - one for the query execution and n for loading each of the entities (not counting additional fetches of references).

This has suboptimal performance. It would be better to be able to read the entities directly from the query results, thus reducing the number of repository calls to 1 (not counting additional reference fetches).