But that means any progression, if it is to be effective, must be used at the right time. How is anyone to know what that right time is if (as you say in no. 1 in your list above) the game is totally random and unpredictable?
Now you asked something wiser. As I said, we can not make any guess and rather need not. Do we need to guess the order of wins in a martingale or labby? We just need a ratio of win. A Marty or labby can beat any session mathematically but they are ill formed to need huge chips . Similarly, I can adopt a progression approach that can make me win if I get a ratio, even after a few thousand spins. Instead of expecting a win in 37 spins, I can handle if I get an average of 40. Instead of predicting what is going to happen in a few next spins, I wait for the average getting closer to my target range. Even if it doesn't go any close, I have ways to sustain the worst without much damage. Gradually, I will win or not loose big enough to recover later.
Do not consider every progression approach to be a push to win 1 unit risking 1000s. Sadly, there has been negligible work in this area.