To share some of my own experiences with TA, and more specifically with indicators:
As an observation I notice that most people (as appears also in the examples above here) use a single timeframe and watch for their signal there - whatever that signal may be. This is trading without looking at the larger timeframe(s) context.
I've read a couple of times this notion that an entry is sort of like a coin flip, having a 50% chance of being right. (whatever "right" is with trading, but that's another story)
Assume you take your long and short signals with a 50% hitrate of a 5-min chart. This will provide winners and loosers. (I don't like the term "hitrate", because I don't believe it's the key to being succesfull at trading, but that's another story too

)
If you would add to your view also a 15-min and an hourly chart, you might notice that if the 15-min and hourly chart are trending up, it might not be a great idea to take those short signals on your 5min chart.
This can already be a first way to get your hitrate up, or decrease the number of losing trades.
Imagine you would jump and watch the soil under your feet. Suddenly you would jump 3 times as far and high. Looking at the soil leaves you clueless how that suddenly happened. Looking at the bigger picture, you might notice you're on the moon instead of earth with different gravity rules.
It's similar with trading. Your entry can have different results depending on the bigger context environment. So loose that keyhole view and add a few timeframes extra and see if behaviour makes more sense to you now.
And then there's the issue of "I have backtested all indicators and none work..."
I wonder if they tested that with different timeframes. But nevertheless, this approach is futile. It's the search to a "get me rich quick" button, but it does not give the trader in any way a feeling with price movement or how the market operates. And I noticed it above with the mentioning of the black swan movemens intraday which were nothing more then "normal" movements after some important news came out.
There are other ways to tackle (get to know) indicators too. You could stick to just one and try to get the most out of it.
Personally I'm a big fan of stochastics in a multitimeframe environment. I use it to determine if there is a trend , what the trend is, what the strength of the trend is. To see if we are grinding up/down. To check if I expect the trend to continue or to reverse etc.
But it required some dedication to stick with this one indicator. That is expressed in months/years not in minutes/hours. So it always gets a smile on my face when somebody says he tested and indicator for a couple of hours, maybe days even and then concludes it doesn't "work"
So don't give up too quick on that. Trading is a job that comes with failure, learning and experience. Just like anything else in life, job or sports or hobby...
That's it for my yearly post here
Xspurt, I must admire you for jumping into this snakepit once more. But I guess if one or two can see the light, it might be worth it.
JP