When configured to ignore orphan contents, web content filters return such contents in getMatchingContents()'s Iterable just the same, and it's the responsibility of the caller to test and filter out the results.

It could be interesting to implement a FilteringAmetysObjectIterable (like commons-collection's FilterIterator or guava's FluentIterable with filter) which would automatically filter out unwanted elements.
The downside is size() would be forced to return -1.