Personally, I think the biggest reason to be excited about SQL Server on Linux is because it makes it viable for much of the Open Source world to start supporting SQL Server as a backend.
For instance, take the Ruby on Rails world. Supporting SQL Server is really difficult in that environment, because the Ruby community takes testing seriously. Most projects won't accept commits that don't include tests. That test suite needs to be automatable, preferably on something like Travis. People check in code to github, that automatically fires up a request to Travis CI, and 10 minutes later the fully automated test suite has run hundreds or thousands of tests on multiple versions of Rails, Ruby, etc. and reports whether the code is good. That model only works for SQL Server when SQL Server can run for free on Travis CI.
I'm not sure what the current state of support is for SQL Server on Travis CI - it's not in their documentation yet (see https://docs.travis-ci.com/user/database-setup/). But it's possible now, and once it happens that will enable a whole universe of robust and tested support for SQL Server in a variety of Open Source development environments.