Start measurement of query time metric w/ actual query

[0snap]

📔 Description

Start the measurement of the query time metric together with the actual indicator. This way, we don't skew the metric with indicators that cannot be processed due to user settings for max-background tasks.

📝 Checklist

• All user-facing changes have changelog entries.
• The changes are reflected on docs.tenzir.com/threatbus, if necessary.
• The PR description contains instructions for the reviewer, if necessary.

🎯 Review Instructions

n/t

[lava]

I kinda assumed it was intentionally outside of the async-block while doing the initial review, so we can see when threatbus is backlogged by its open tasks.

But, either way is fine with me.

[0snap]

We now have a dedicated metric to track said backlog, so this one can go into the async block 👍