Resolve TMSS-337

Closes TMSS-337

Code looks good, can be merged. Little confused with the status properties on task and scheduling unit and why the implementation differs from explicitly modeled subtask state, though. But yeah, that was not introduced in this ticket.

Actually, I think I understand why they are not modeled (e.g. because a FK reference might clash with subtask states and we always want to derive). But nonetheless, I think we should have a reference table on what states are to be expected here and maybe check that we don't return 'forbidden' values? Anyway, I guess I was just confused because 'observed_end_time' had no intuitive meaning to me... ;)

resolved all threads

merged

mentioned in commit 67d7543c