more thorough testing of agent parameter on GET Statements requests

[cr8onski]

The current tests in the test suite retrieve statements from an LRS using the agent parameter and a specific agent, then test that the requested agent is contained in the retrieved statements.

The above testing still leaves a hole where an LRS could pass that test, however not be able to meet a case where a just a specific IFI is used to query an LRS.

According to my reading of the specification, an LRS should return any statement with an actor or object which has an agent or group which contains that IFI, regardless of other agent or group information. Communication 2.1.3 Table row 3

If this is right, I propose the new tests to cover this gap. Also would this need an additional requirement?