Comparing complex projects is not really fair. One project maybe slow when used end-to-end, while the other is blazing fast. Sure there will be a difference in how fast your feature tests will run.
- Run your tests in parallel in the cloud (Fast but costly)
- Profile the tests and your application code
- Analyse double coverage and try to move tests lower on the test-pyramide if possible
Measuring and researching is knowing. First analyse the slowest tests, execute it manually and check if the slowness is the application or the test. Optimize accordingly.
If you cannot optimize, run everything in parallel. If you build your tests to run in isolation this should be easy.