envoy-gateway only binding v4 addresses on dual stack

*Description*:

I have a dual stack cluster with v6 as the preferential IP stack (i.e. in all configs that take multiple IP addresses, the v6 address appears first). This means that everything is done preferentially via ipv6. The kubelet performs health checks via the pod's v6 address and not v4 address.

During gateway startup, envoy-gateway is configuring envoy to bind 0.0.0.0 for a load of ports (such as healthcheck on 19003) which only listens on v4. This means on my cluster the pod can never become ready because the kubelet is trying the pod's v6 address for the healthcheck which is not available.

It seems like this is a similar issue to https://github.com/envoyproxy/gateway/issues/6389.