Reordering HTTPRoute rules causes brief 503 errors to clients

*Description*:
I believe this is similar to (or the same as) #6287, but I'm opening a new issue since we hit this problem in an actual production use case.

For readability, we had been defining the **last** rule in `HTTPRoute.spec.rules` as a catch-all default rule (`path: { type: PathPrefix, value: / }`), and placing newly added rules above it in whatever order was easiest to read.

However, with Envoy, whenever the **index of a rule changes**, clients receive `503` responses for several seconds.

*Repro steps*:

Switching back and forth between the two manifests below triggers `503` errors **every time**. The only meaningful difference is that the `/` PathPrefix rule moves from index `0` (Example A) to index `1` (Example B).

**Example A:**

```yaml
apiVersion: gateway.networking.k8s.io/v1
kind: HTTPRoute
metadata:
name: site
spec:
parentRefs:
- name: eg
hostnames: ["example.com"]
rules:
- matches:
- { path: { type: PathPrefix, value: / } }
backendRefs:
- { name: frontend, port: 80 }
```

**Example B:**

```yaml
apiVersion: gateway.networking.k8s.io/v1
kind: HTTPRoute
metadata:
name: site
spec:
parentRefs:
- name: eg
hostnames: ["example.com"]
rules:
- matches:
- { path: { type: PathPrefix, value: /api/ } }
backendRefs:
- { name: backend, port: 80 }
- matches:
- { path: { type: PathPrefix, value: / } }
backendRefs:
- { name: frontend, port: 80 }
```

Applying A, then B, then A (etc.) produces a short window of `503` responses on each transition.

*Environment*:
Envoy Gateway version: `1.8.0`