Feat: Better reporting if something is wrong with taskprocessing

[marcelklehr]

Summary

It turns out that the occ taskprocessing:worker command does not run when the instance is in maintenance mode and then usually just stops (after a few failed restarts). This adds a setup check to make the admin aware that the worker is not running, iff there have been scheduled tasks in the last 7 days (to make sure we don't warn if the feature is not used).
The second commit increases the sensitivity threshold for the other taskprocessing setup checks from 20% to 10% to make them warn earlier.

Checklist

• Code is properly formatted
• Sign-off message is added to all commits
• Tests (unit, integration, api and/or acceptance) are included
• Screenshots before/after for front-end changes
• Documentation (manuals or wiki) has been updated or is not required
• Backports requested where applicable (ex: critical bugfixes)
• Labels added where applicable (ex: bug/enhancement, 3. to review, feature component)
• Milestone added for target branch/version (ex: 32.x for stable32)

AI (if applicable)

• The content of this PR was partly or fully generated using AI

[julien-nc]

I don't see the check results. I tried with no tasks recently scheduled and after having scheduled one.
There should always be one, right?

[marcelklehr]

I don't see the check results.

Ah, it's not registered.

[oleksandr-nc]

Looks good!

[marcelklehr]

/backport to stable33

[marcelklehr]

/backport to stable32