Why only send one notification object at a time?

We can specify send interval in config options, but even so at that interval we should send all unsent notifications in the queue each time instead of just one.

Is there a performance reason that you are only sending one notification object per loop?

The issue does not name files, tests, or an entry point. Locate the notification queue and its send loop, determine whether sending all unsent notifications at each configured interval is safe, and document the performance implications; done means the behavior and its coverage are clear.

Written by the indexing model from the issue text.