I'm pretty sure that the optimal f calculation assumes that your worst loss is still in the future, using that information to determine bet size between now and the occurrence of that loss to reduce risk of ruin or maximum drawdown. To the OP, if you are just going through a phase where the system does not work, you may want to shut it down until it starts to work again and optimal f goes positive, rather than throw it away. That's a judgment call.
If you purchased a copy of this book you should return it and ask for a refund. The authors don't know what they are talking about. They are totally confused in this subject. Their formula is totally wrong. I wonder how many traders they have misguided. The correct formula gives 68.15% for optimal %Kelly rather than 79.2% for the example in that book excerpt.
Sorry, but how did you get 68.15%? I plugged the numbers in w - (1-w)/R and I got 0.7929. I think the main discrepancy between the formula you gave me and my book is using the payoff ratio as opposed to the profit factor. I suspect that my book is wrong because I have another book that uses the payoff ratio.