JobSubscriberIntakeQueue Cleanup

[adeutscher]

A few not-on-fire cleanup items for JobSubscriberIntakeQueue that I noticed while working on another issue. Marking the ToDo here for branch blast radius reasons.

• Add some comments, especially to confirm our thinking around our cancellation strategy.
• Revisit why we suppressed Sonar's S2325 around Load. At the very least, correct the disable/restore thing.
• Consider moving the reset of the event to south of the null.

[adeutscher]

PR created. Will do local testing before merging tomorrow.

On revision, I have no idea why I was suppressing S2325 in the first place. S2325 is the 'this method can be made static' one, and it was on the interface implementations. It's not like I can blame Cursor for this, all my Sonar suppression is manual. Weird.

[adeutscher]

Local testing looks good. I probably made the intake queue a bit more threadsafe than it really needs to be, but I'm happy with it.