Determine process to ensure successful alpha upgrade

Needless to say, there's a ton at stake in the upgrade process. What can/should we do to ensure that there are absolutely no mistakes?

Some ideas we've considered:
* In other complex, high-stakes environments (e.g. aerospace, surgery, construction) checklists are used to make sure important steps don't get missed. Off the cuff, I can think of quite a few things that we definitely don't want to forget to do before running the upgrade script, e.g.:
- [ ] correct ENV is loaded in shell
- [ ] there is a co-pilot to watch and sanity-check everything being done
- [ ] we're using the correct forge version
- [ ] all tests pass locally
- [ ] all tests pass on CI
- [ ] we're on the correct git commit locally
- [ ] dry-run succeeds
- [ ] test run on local fork looks as expected
- [ ] other devs are around to observe/triage any issues
- [ ] we have open channels with important stakeholders
- [ ] then, only after all that is done, run the upgrade script live
* Build a fail-safe into the proposal in some way, e.g. that a certain important function can only be executed if `tx.origin` is an address of our choosing
* Write new fork tests against mainnet at each stage of the process
* Make the transfer a two step process with two separate proposals, e.g. one to the AlphaGovernor to update the Timelock admin, a second to the BravoGovernor to accept the admin, so that the upgrade can't happen unless proposals really can be voted on and executed by Bravo