Support test case sub-results per performance variable

[vkarak]

In a test with multiple performance variables (figures of merit), if one of them fails in performance, the whole tests is marked (correctly) as a failure. However, when logging the individual sub-results, there is no way to differentiate a success or failure and this can be confusing. It would be desirable to introduce a "sub-result" concept per figure of merit.