Find a way to identify combined test executions

At the moment each test result has a unique ID, but it's not possible to determine when two test results belonged to a single execution of a test suite.

It's not possible to determine whether e.g. `ltp.stat01` and `ltp.stat02` belong to the same execution of `ltp` or to two different ones. It is therefore impossible to know how many executions of `ltp` there actually was. We can only find either the minimum, or the maximum possible number of executions, but not the actual number. It is also impossible to determine how many particular outcomes was there, as it's not possible to attribute a particular test to a particular suite and calculate its independent outcome.

Find out if this would be good to know, and if yes, how to implement identifying particular executions.

The issue names no files, tests, or entry points, so first inspect the repository's data model and test-result handling to understand how executions are represented. Confirm whether grouping results into suite executions is needed, then document an implementation approach and the data needed to produce per-execution counts and outcomes.

Written by the indexing model from the issue text.