If you apply the concept to markets blindly, that's true. If you use a little creativity in its application, its still a pretty effective method of trend capture.
Depends why you entered. If you are entering because you expect to catch the next leg of upward momentum, then sell once the momentum is no longer there. You may well find that 50% of the time (or more), the momentum resumes later. That doesn't mean it was a mistake to sell when the initial momentum ended. If you are entering to catch a major trend, then you should not sell until there are clear signs the trend is at an end. It is pretty much impossible to ride a multi month or multi-year trend without experiencing some pullbacks. But it IS possible to ride a stock up 200%+, or ride a futures market up 100%+, using trend-following techniques with a wide trailing stop.
I have read some good ideas on ET about having some type of go/no go context filters for general market conditions that would turn off or moderate a breakout trend following system in light of present general conditions. Don't remember specifically where , but, probably in the Strategy System design section. Good luck.