The figures you mentioned earlier are also based on the personal observations of an unknown author and published in a physics magazine. I don't think you verified them either, but you still post them for everyone to see, presenting them as truth.

But that's not the case. The morale works as follows. Let's say we have a morale of +3. At the moment of attack (movement), a random number is chosen. This number is divided by 24. The remainder of the division can be 0, 1, 2... 23. If the remainder is 1, 2, or 3, then the morale takes effect; if not, it doesn't. Perhaps it's more complex than that, but that's how I understand it.
3 out of 24 is 12.5%.
2 out of 24 is 8.33%.
1 out of 24 is 4.2%.
Based on this, it's possible to achieve a situation where morale decreases with each turn.
This data was obtained by examining the exe file, so I believe the data that AlexSpl provided, but I'm not very confident in personal observations and some tests. You can test flipping a coin and come to the conclusion that the probability is not 50/50.