[BUG] V6.0 pushes API with empty serviceUrl

### Release version

v6.0.11

### Describe the bug

APIOps removes the `serviceUrl` (also known as HTTP(s) Endpoint) configuration for some of our APIs.
the serviceUrl is overwritten for every environment in a configuration and a couple (that we have seen) of our APIs had the serviceUrl changed to `null` by APIOps after upgrading to v6.0.11

I just rolled back to v5 and the few API have got their serviceUrl back

### Expected behavior

API consistently deployed with serviceUrl if it is overwritten from configuration file

### Actual behavior

few API deploy with serviceUrl as `null`

### Reproduction Steps

The apiInformation, specs and config file did not change since v5 for those API so I am unsure why it would fail here.

apiInformation.json:

```json
{
"properties": {
"apiRevision": "1",
"authenticationSettings": {},
"displayName": "rp-sis",
"isCurrent": true,
"path": "rp/scheme-integration-service",
"protocols": ["https"],
"serviceUrl": "https://sis.example.net",
"subscriptionRequired": false
}
}
```

configuration.d1.yml:

```yaml
---
apimServiceName: apimxxxxx
[...]
apis:
[...]
- name: rp-sis
properties:
serviceUrl: "https://sis.dev-example.net"
[...]
```