Remaining todos:

• should we support favorPerFieldSimilarity (see discussion on mailing list and PR related to DisMax)
• genericize the DSL to receive the entity type for a typesafe .toEntity(...)
• improve FieldCustomization interface to split FieldBridge from Analyzer options as fieldbridge are ignored by MLT
• should we support toContent(Reader) and toContent(String): I did not find a use case
• does the TypeMetadata return the embedded properties (question asked to Hardy Ferentschik during IRC meeting)
• should we add additional stop word input. I feel it could be useful but need to refine the use case
• what to do about min/maxTermFreq min/maxWordLen maxTokenParsed maxQueryTerms? Should we expose them?
• support numbers with exact match
• support compressed data
• beef up test suite with complex entities with associations, number fields, custom fieldbridges, class bridges etc