fix: Apply schema and record transformations for v2 sources with v3 destinations (#13291)

#### Summary

Fixes https://github.com/justmiles/cq-source-crowdstrike/issues/5.
We should apply the schema and record transformations when the source is v2 and the destination is v3, otherwise pks are not removed in `append` mode, sync time and source name columns are not added, etc.

**This probably impacts all community/partners plugins as I believe those haven't upgraded to SDK v4.**

Tried to do this the less possible ugly way I could think of.
Tested with:

```yaml
kind: source
spec:
  name: azure
  path: cloudquery/azure
  version: v8.3.0 # v2 source
  destinations: ["bigquery", "postgresql"]
  tables: ["azure_storage_*"]
---
kind: destination
spec:
  name: bigquery
  path: cloudquery/bigquery
  version: "v3.3.0" # v3 destination
  write_mode: "append"
  spec:
    project_id: ****
    dataset_id: ****
---
kind: destination
spec:
  name: "postgresql"
  registry: "github"
  path: "cloudquery/postgresql"
  version: "v4.2.2" # v2 destination
  spec:
    connection_string: ****
```

<!--