If you have multiple people on your team, a common dev environment can be a huge time-saver. New developers can install an identical environment quickly.
Another benefit is that your project's development environment is self-contained. It's not going to clash with anything else on your system, and will even work across platforms.
You can do a pip install of python bindings but not the source package itself. For e.g., pip install redis gets you python bindings for the redis but not redis itself. With this setup, you can install a specific version of redis, run it as a service with your custom configuration right out of the box.
I'm guessing easy resource allocation, either through local machine quotas/cgroups or the easier ability to move specific services to different machines as you need to scale them.
Another benefit is that your project's development environment is self-contained. It's not going to clash with anything else on your system, and will even work across platforms.