jobs,bulk,*: Provide more visibility into job status

[dt]

Better per-processor status reporting -- what is throughput, where is it blocked, what is recent throughput, are we encountering underlying errors / retrying / etc. More visibility into what jobs are doing in general. TODO(dt, bulk): brainstorm and write out some specific sub-goals here.

[mwang1026]

• Debugging endpoint--be able to get there and see status of every goroutine and processor working on behalf of this job
• Ideally throughput on processor and node level--immediately see where throughput is lowest
• Revising milestones as we start using this more

---

• (?) Job progress table?

---

• Getting onto jobs page is a bit more work

[mwang1026]

PRs coming soon

[livlobo]

Closing in favor of other tracing work here: #64992