If Gitpod (!) or GitHub Codespaces is down (!) then if your team has successfully migrated towards reproducible environments then mitigation in disaster scenarios should be as simple as
So if our workspace environments are fully reproducible and trivial to deploy already, what additional benefits do these services bring to the table? (circling back to my original question :) )
Been using internally-managed reproducible environments for a while now, so I'm fully aware of the major benefits.
The code is still in git and on GitHub, the workspace is ephemeral. You can always clone the repo and work locally. I think the idea isn't to 100% replace all local workflows necessarily but to supplement them and to enable working in situations where you don't have access to a developer workstation otherwise.
Of course if you roll this out to the entire company to replace regular workstations, you probably want something with an SLA. Not sure what Gitpod's answer is, but Microsoft seems to sell Codespaces outside GitHub via Azure and there's probably a story for the enterprise version addressing this need.
I don't see a way of committing to use this service without also maintaining the capability required to operate as you are now on a moments notice. The SLAs won't be cheap either I imagine.
Don't get me wrong though, I think it's a cool service and I can see it going places. It just seems impossible to justify for small to medium sized shops, especially considering you can get the exact same benefits from containers and images right now (and even run your own centralised VSCode server).