Fix error branches, take 3.

[riastradh-brave]

• Make sure we continue watching for tor restart after an error on the control channel.
• If we get a watch event while polling, don't cause us to spin polling as fast as the CPU can go until tor is launched.
• Don't duplicate errors synchronously and asynchronously.
• Handle asynchronous errors gracefully.
• Fix some tor redundancy in tor log messages about tor that say tor a few too many times.

fix #14630 for real, maybe?

Submitter Checklist:

• Submitted a ticket for my issue if one did not already exist.
• Used Github auto-closing keywords in the commit message.
• Added/updated tests for this change (for new code or code which already has tests).
• Ran git rebase -i to squash commits (if needed).
• Tagged reviewers and labelled the pull request as needed.
• Request a security/privacy review as needed. (Ask a Brave employee to help if you cannot access this document.)

Test Plan:

1. Turn off the network.
2. Launch Brave.
3. Open a private tab with Tor.
4. Wait for it report timeout.
5. Hit the 'Retry' button as fast as your fingers can manage.
6. Confirm no stack traces dumped to the console.
7. After a minute, turn the network back on.
8. Visit https://check.torproject.org to confirm Tor works.

Reviewer Checklist:

• Request a security/privacy review as needed if one was not already requested.

Tests

• Adequate test coverage exists to prevent regressions
• Tests should be independent and work correctly when run individually or as a suite ref
• New files have MPL2 license header

[bsclifton]

Tested fairly exhaustively on Windows 10 and these changes worked great for me 👍

[bsclifton]

master 684ed1f
0.24.x 764eb6e
0.23.x f456bac