librarian: when designing repo consolidation, where should samples go?

We have *at least* 4 possible kinds of sample code. (I know of 4 for .NET...)

- Product-written code, typically in a separate language-specific repo, e.g. http://github.com/GoogleCloudPlatform/dotnet-docs-samples; these are included in product docs
- Autogenerated monolithic snippets, e.g. https://github.com/googleapis/google-cloud-dotnet/tree/main/apis/Google.Cloud.SecretManager.V1/Google.Cloud.SecretManager.V1.Snippets
- Autogenerated file-per-method snippets, e.g. https://github.com/googleapis/google-cloud-dotnet/tree/main/apis/Google.Cloud.SecretManager.V1/Google.Cloud.SecretManager.V1.GeneratedSnippets
- Handwritten language-specific code included in our reference docs, e.g. https://github.com/googleapis/google-cloud-dotnet/tree/main/apis/Google.Cloud.Firestore/Google.Cloud.Firestore.Snippets which includes code shown in https://docs.cloud.google.com/dotnet/docs/reference/Google.Cloud.Firestore/latest/datamodel; in .NET these are written in the form of integration tests

Some languages may have other kinds of sample code, or not have as many kinds as this.

We need to work out where it will all live in the future, and whether we want it all.

Start by inventorying the four sample locations described in the issue, including dotnet-docs-samples and the SecretManager and Firestore snippet paths. Compare how each language handles product, generated, and reference samples; done means a documented decision about which categories to retain and where they should live.

Written by the indexing model from the issue text.