Determine what, if any, additional tests can be defined

Currently, the `Executable Specifications` entry in `imm.yaml` scores just 1 point. There are two additional ways we can increase the score:

* Specs which cover common failure cases
* Specs which explore edge cases

We can do either (for one point each) or both (for two). Arguably the tests already cover edge cases as far as possible, but someone other than @idg10 should review them to see what opportunities I have missed to cover more.

We don't currently have much by way of testing for failure, not least because we haven't really defined what we want to have happen. Failure will generally mean that the underlying instrumentation provider is throwing exceptions. Probably all we can do in this case is allow them to pass back out unimpeded. This is likely to be the default behaviour already, but we could add tests to verify that.

Start by reviewing the `Executable Specifications` entry in `imm.yaml` and the existing tests that contribute to its score. Look for missed common failure cases and edge cases, and determine whether instrumentation-provider exceptions should pass through unchanged. Done means adding justified coverage where opportunities exist and verifying the resulting score.

Written by the indexing model from the issue text.