I have read what is written here and I wanted to say there a lot of good interesting points. I can’t respond to them all do them all justice. Let me say two things.
First this Zorro thing is a special resource: the manual is incredibly comprehensive on how to write a full featured trading system yourself. I think the sample strategies and published expected returns are trying to answer the first question you ask. They are saying: yes it is possible - here you go - here is a trading system/strategy that works - here is a complete how-to guide on how to write one – now go do it. To believe what they claim or not is up to you though.
Second ‘what is the best way to find a strategy’? Well how did Zorro find their strategies. In the manual they mention a book by David Elhers. I’ve not read it but a lot of these functions they use like Lowpass and Fisher etc come from the book. My guess is the strategies are outlined in there too. So they read his book, implemented the strategies in Lite-c, they worked out the strategy parameters through ML/Optimization and there you go. This is quite common in Physical sciences. Academics publish research papers and practitioners implement them. So to get strategies you have to read – papers – blogs - books etc. I’ve seen lots of good lists of resources published so I won’t list them. ‘Trading System and Methods by Kaufman’ is 1200 page tomb I’ve read and heard called the bible of strategies.
I would say I have read a trading paper a day or a book a week for 4 straight years – and yes I prefer this to video games! J