Hi, thanks for sharing this issue.
Your configuration looks correct, so the first thing that comes to my mind is your .cache file.
When NCrunch runs your tests for the first time, it has no knowledge of how long each test takes to execute. This means that it cannot create an efficient pipeline and it will often struggle to execute small suites in parallel. Once the first run through has completed, NCrunch will save the test execution times inside a .cache file under a '_NCrunch_solution_v2' directory next to your .sln file.
This is an important consideration for when you are running the console tool inside a CI system, as CI systems will often perform a fresh check-out of the source code (without the .cache file) when they perform a build. If you don't have this .cache file present and you only have 15 test cases, I would say it's likely your parallel execution won't be working efficiently.
Recently, a new setting was introduced that allows you to control where NCrunch will store its .cache file (
http://www.ncrunch.net/documentation/reference_solution-configuration_ncrunch-cache-storage-path). I'd suggest setting this to an absolute path on the CI server so that the .cache file always exists in the same place and can be accessed from any location the CI checks out the code.