fix(NODE-6845): ensure internal rejections are handled

[nbbeeken]

Description

What is changing?

• we convert unhandled rejections to uncaught exceptions

Is there new documentation needed for these changes?

No

What is the motivation for this change?

Recent issues with KMS and unhandled rejections inspired better defaults for our test suite

Fix misc unhandled rejections under special conditions

We identified an issue with our test suite that suppressed catching unhandled rejections and surfacing them to us so we can ensure the driver handles any possible rejections. Luckily only 3 cases were identified and each was under a flagged or specialized code path that may not have been in use:

• If the MongoClient was configured to use OIDC and an AbortSignal was aborted on cursor at the same time the client was reauthenticating, if the reauth process was rejected it would have been unhandled.
• If timeoutMS was used and the timeout expired before an operation reached the server selection step the operation would throw the expected timeout error but a promise representing the timeout would also raise an unhandled rejection.
• If a change stream was closed while processing a change event it was possible for the "change stream is closed" error to be emitted as an error event and reject an internal promise representing fetching the "next" change.

Double check the following

• Ran npm run check:lint script
• Self-review completed using the steps outlined here
• PR title follows the correct format: type(NODE-xxxx)[!]: description
  • Example: feat(NODE-1234)!: rewriting everything in coffeescript
• Changes are covered by tests
• New TODOs have a related JIRA ticket