Note1: Some of the validation steps in this workflow are based on statistic/probabilities, if you eventually will filter out some good strategies, a good practice is to go back to test 1 (OOS1) and run those strategies through the whole process again.
I am sorry, I did not understand this statement. Can you re-explain it? Thanks.
The MC tests are done using random test parameter generation, if you will use a very low number of MC runs (e.g.: 100 runs, to speed up the processing time in your workflow), it can happen that a given strategy will pass this test only due to luck and not because it is a robust strategy. In that case you need to run MC tests with a much larger number of runs or repeat the same test few times to see if your strategy really passes this test each single time.