You’re confusing “parallel” from an audio engineer’s point of view with what is computationally parallel. What you call parallel is actually computationally serial. Anyway, a better test would be to use the same conditions across DAWs instead of different conditions. And the best metric of performance is always maximum number of tracks, not % of cpu capacity used.
Anyway, the purpose of this thread is to provide a benchmark project file, included in the first post, that others can download and try for themselves. If you have test results for that benchmark, please post your results here. If you want to discuss a different test with your own benchmark, it would better to start a new thread, rather than confuse the issue here with results posted from multiple benchmarks.