Deploy a single data model from a repository with multiple models defined

This is not possible. If multiple models are defined within a single repository, which is possible, you are unable to target a specific model within that repository for deployment. The deployment process is at the repository level. 

This can be effectively navigated by adopting industry best practices for proper git repo and branch structuring.

The impact of this behavior is different across environments.

• In PROD, we don’t expect any issues. Data models promoted to a production environment will only be active data models and therefore deploying (making query available) all models would be expected.
• In UAT, we don’t expect many issues. Good user acceptance testing practice would be to mimic a PROD environment and only have data models that are actively tested. As such, deployment of all models should not cause issues. If during UAT a data model is deemed ‘not ready’, it can be deleted from a prod (main) branch awaiting dev work and promotion to the UAT branch.
• In DEV, it could become cumbersome as a development team might have many projects in various stages of development. The customer can introduce some simple Git practices that alleviate issues. Allowing for personal branches would allow a developer to work on their models, deploy from personal branch and only on success promote it to a higher branch - such as a collective DEV.