Agents which fail creation are sometime added to the node list and never removed

### Jenkins and plugins versions report

Environment

```text
Google Compute Engine Plugin Version 4.815.v9a_de97053e3a_
```

### What Operating System are you using (both controller, and any agents involved in the problem)?

Ubuntu

### Reproduction steps

1. Configure a agent type with fallback zones and fallback configurations which depend on limited resources such as GPUs
2. Create instances that lack resources

### Expected Results

Any agent that does not exist in GCP is either never added to the list of agents in jenkins or gets removed after a timeout

### Actual Results

Some agents get added to jenkins despite not existing in gcp, which will prevent functional agent being created because jenkins thinks there are already enough.

### Anything else?

The fallback system does work including fallback configurations that create agents in other zones, though some of the agents created with the fallback configuration also fail to create correctly.
Looking at the logs I can't see what is different between creation attempts that correctly fallback and the ones that leave a dead agent.

### Are you interested in contributing a fix?

I can get time to work on this issue, but I have little knowledge of the plugin so would need guidance identifying the root cause.

No source files or tests are named. Start by reproducing the fallback-zone and limited-resource scenario from the issue, then compare Jenkins and GCP state with the available logs. Done means failed creations are not retained as Jenkins agents or are removed after a timeout, while fallback configurations can still create functional agents.

Written by the indexing model from the issue text.