feat: Add EKF odometry baseline example

[djetshu]

Proposed changes

Adds examples/ekf/, a new worked example alongside examples/beluga/. It sweeps a single algorithm variant (raw_odom vs ekf_fused) to compare open-loop wheel odometry against a robot_localization EKF fusion, scored with evo_rpe (drift-rate, since open-loop odometry drifts unboundedly). Ships its own ROS 2 package and Docker/Docker Compose environment, reusing the same bundled dataset image as Beluga.

Type of change

• 🐛 Bugfix (change which fixes an issue)
• 🚀 Feature (change which adds functionality)
• 📚 Documentation (change which fixes or extends documentation)1

💥 Breaking change! Explain why a non-backwards compatible change is necessary or remove this line entirely if not applicable.

Checklist

Put an x in the boxes that apply. This is simply a reminder of what we will require before merging your code.

• Lint and unit tests (if any) pass locally with my changes
• I have added tests that prove my fix is effective or that my feature works
• I have added necessary documentation (if appropriate)

Additional comments

• No automated tests, same as examples/beluga/ — validated by running the benchmark end-to-end in the dev container.
• Docker only; Podman support left out since it wasn't verified.
• Worth flagging for review: nominal() deliberately skips --clock on ros2 bag play, since the dataset already carries its own recorded /clock topic. Passing it too starts a conflicting second publisher that corrupts ekf_node's sim time and silently truncates the trajectory — see the README's "Sim time and the recorded /clock topic" section.