Seems like a question for Professor Duncan or Cosima.

It really depends on which Socket Services the container requires. If you have a lot of containers that all need the same set of Socket Services, you could potentially use a single socket-proxy to serve all of them (in theory, I think).

I usually run one per stack, sometimes more if I have a container within my stack that requires more/different Socket Services to the other(s).

I’m not a docker expert though, so I’m not sure I can say what’s recommended. If you find/get a more authoritative answer on this question, I’d be interested to know.

The risk is certainly lower if you’re not exposing services to the internet, but that’s not the only way to end up with a rogue container. I use docker-socket-proxy for most of my stacks that need socket access. It can sometimes require a little bit of troubleshooting to understand what services you need to proxy, but I’ve had a pretty good success rate. Reading the logs from the socket-proxy and referring to the Docker Engine API documentation will help you to understand what Services you need to enable in the socket-proxy config.

Are you interested in the networking side of self hosting? If so, you should get a better router, something you can run OPNsense or similar on. There are other “options”, but they’re workarounds that avoid fixing the real problem.

They don’t want people installing GrapheneOS on them.