### Brief description of Pull Request

This PR prevents a panic case where we create multiple loaders for the
same controller. We can identify when we have created duplicate loaders
when the prometheus metrics owned by the loader already exist.
Attempting the register the metrics a second time resulted in the panic.

### Issue(s) fixed by this Pull Request

Fixes: #2801

### Notes to the Reviewer

This issue can be reproduced consistently by,

1. Run alloy in docker with minimal resources + remotecfg
2. Create a pipeline in fleet management which uses a component that
bind a resource on Build (ex loki.source.api binds a port)
3. Rename the pipeline in fleet management
4. The new config fails to load because the port is already bound
5. The remotecfg component attempts to load the old config
6. Duplicate loader is created resulting in panic

Abridged logs for what will happen now,
```
alloy-1  | ts=2026-02-04T19:23:55.07699636Z level=debug msg="fetching remote configuration" service=remotecfg
alloy-1  | ts=2026-02-04T19:23:55.202305809Z level=info msg="attempting to parse and load new remote configuration" service=remotecfg config_hash=bace1e80
alloy-1  | ts=2026-02-04T19:23:55.34085171Z level=info msg="finished node evaluation" controller_path=/remotecfg controller_id=source_api.default trace_id=166fa51994d6f7723fe329ca53ea9b9a node_id=loki.write.grafana_cloud_loki duration=92.039756ms
alloy-1  | ts=2026-02-04T19:23:55.345658677Z level=info msg="starting push API server" component_path=/remotecfg/source_api.default component_id=loki.source.api.loki_push_api
alloy-1  | ts=2026-02-04T19:23:55.345713885Z level=info msg="starting server" component_path=/remotecfg/source_api.default component_id=loki.source.api.loki_push_api
alloy-1  | ts=2026-02-04T19:23:55.349334402Z level=error msg="failed to evaluate config" controller_path=/remotecfg controller_id=source_api.default trace_id=166fa51994d6f7723fe329ca53ea9b9a node=loki.source.api.loki_push_api err="building component: failed to run embedded server: listen tcp 0.0.0.0:9999: bind: address already in use"
alloy-1  | ts=2026-02-04T19:23:55.349425319Z level=info msg="finished node evaluation" controller_path=/remotecfg controller_id=source_api.default trace_id=166fa51994d6f7723fe329ca53ea9b9a node_id=loki.source.api.loki_push_api duration=8.420068ms
alloy-1  | ts=2026-02-04T19:23:55.349562734Z level=error msg="failed to evaluate config" controller_path=/ controller_id=remotecfg trace_id=747e92eb01b59f96fe555b578fc9fb6c node=source_api.default err="updating custom component: 129:2: Failed to build component: building component: failed to run embedded server: listen tcp 0.0.0.0:9999: bind: address already in use"
alloy-1  | ts=2026-02-04T19:23:55.349575859Z level=info msg="finished node evaluation" controller_path=/ controller_id=remotecfg trace_id=747e92eb01b59f96fe555b578fc9fb6c node_id=source_api.default duration=101.567234ms
alloy-1  | ts=2026-02-04T19:23:55.353856414Z level=error msg="failed to parse and load configuration" service=remotecfg config_size=8280 err="129:2: Failed to build component: building component: failed to run embedded server: listen tcp 0.0.0.0:9999: bind: address already in use"
alloy-1  | ts=2026-02-04T19:23:55.353934996Z level=error msg="failed to parse and load new remote configuration" service=remotecfg received_hash=bace1e80 loaded_hash=775b38f6 err="129:2: Failed to build component: building component: failed to run embedded server: listen tcp 0.0.0.0:9999: bind: address already in use"
alloy-1  | ts=2026-02-04T19:23:55.353956746Z level=info msg="attempting to reload cached configuration to restore component health" service=remotecfg
alloy-1  | ts=2026-02-04T19:23:55.354616409Z level=info msg="finished node evaluation" controller_path=/ controller_id=remotecfg trace_id=cd603bde9ef19de4c6bc63e6f03685ed node_id=declare.source_api_rename duration=2.5µs
alloy-1  | ts=2026-02-04T19:23:55.446986412Z level=error msg="failed to evaluate config" controller_path=/ controller_id=remotecfg trace_id=cd603bde9ef19de4c6bc63e6f03685ed node=source_api_rename.default err="creating custom component controller: failed to create module controller: failed to build loader: a loader exists already exists for \"remotecfg/source_api_rename.default\""
alloy-1  | ts=2026-02-04T19:23:55.447288993Z level=info msg="finished node evaluation" controller_path=/ controller_id=remotecfg trace_id=cd603bde9ef19de4c6bc63e6f03685ed node_id=source_api_rename.default duration=92.65121ms
alloy-1  | ts=2026-02-04T19:23:55.450888052Z level=error msg="failed to parse and load configuration" service=remotecfg config_size=8294 err="157:1: Failed to build component: creating custom component controller: failed to create module controller: failed to build loader: a loader exists already exists for \"remotecfg/source_api_rename.default\""
alloy-1  | ts=2026-02-04T19:23:55.452185752Z level=error msg="failed to reload cached configuration" service=remotecfg err="157:1: Failed to build component: creating custom component controller: failed to create module controller: failed to build loader: a loader exists already exists for \"remotecfg/source_api_rename.default\""
alloy-1  | ts=2026-02-04T19:23:55.45219946Z level=error msg="failed to fetch remote config, continuing with current config" service=remotecfg err="157:1: Failed to build component: creating custom component controller: failed to create module controller: failed to build loader: a loader exists already exists for \"remotecfg/source_api_rename.default\""
alloy-1  | ts=2026-02-04T19:23:55.452297418Z level=debug msg="making immediate GetConfig call to report status update" service=remotecfg

```

The PR also includes some small changes to add more support for
structured logging in the controller.

### PR Checklist

- [x] Tests updated