[uss_quafilier] Lower severity of all cleanup steps

[the-glu]

This PR finalize #1400 by removing checks that where still too high in cleanup phases.

Most of them used fragments used in non-cleanup phrase. I duplicated them in specific _during_cleanup files, but I'm not sure that the best solution.

This has been tested by running whole CI with something like

        if self._phase == ScenarioPhase.CleaningUp and check_documentation.severity in [
            Severity.High,
            Severity.Critical,
        ]:
            raise RuntimeError()

in monitoring/uss_qualifier/scenarios/scenario.py's check() function, showing no left over checks.

[mickmis]

In order to ensure this won't be an issue, what about checking whether we are in the cleanup phase (+ a log entry to have a chance of cleaning that up) here?:

monitoring/monitoring/uss_qualifier/scenarios/scenario.py

Lines 170 to 173 in f9f43e5

	if severity == Severity.High:
	raise ScenarioCannotContinueError(f"{severity}-severity issue: {summary}")
	if severity == Severity.Critical:
	raise TestRunCannotContinueError(f"{severity}-severity issue: {summary}")

(in another PR)

Otherwise LGTM, even though duplication is a bit of y pity, but hard to avoid here.

[the-glu]

In order to ensure this won't be an issue, what about checking whether we are in the cleanup phase (+ a log entry to have a chance of cleaning that up) here?:

Yes, but do we want to enforce that? Someone could decide to write a Cleanup check with High severity.

Also should we choose to enforce that, we could also automatically lower severity in cleanup phase (and remove duplicated fragments)

@BenjaminPelletier any opinion?