I like Japanese candlestick patterns for many reasons. Especially one- or two-line patterns as they appear very often on the charts. Hence, the statistics may be more reliable.
The biggest advantage of Japanese candlestick patterns is their simplicity.
Of course patterns should not be seen as a complete trading strategy in itself. Apart from a decent entry/exit strategy (which makes you a positive expectancy, i.e. you are profitable) you need to have even more decent risk and position management.
If you may find a market edge in a trading strategy using, for example, a 2-line pattern, and if you eliminate to the minimum other parameters, then you have a chance to be profitable. Of course, it does not mean that we can grab any 2-line pattern and think that we have done our homework. But patterns may be great for a bunch of ideas how to enter/exit the market.
I'm writing here more from the position of mechanical trading systems rather than discretionary trading (although patterns may be successfully applied by both groups of traders). And there I prefer a simple entry/exit strategy, which eliminates the risk of curve-fitting. For this purpose, candlestick patterns are really great. If you want to have some complexity, put it in your position sizing to squeeze as much as you can from your market edge. But avoid making entry/exit strategy too complex.
I don't like to "optimize" and make more profits (in backtesting) by applying oscillators and filters. If I know that 2+2=4 (i.e. a pattern appeared), why should I bother by calculation of some derivatives of prices (i.e. oscillators), to have more or less the same result at the end plus the risk of over-optimization? Applying too many filters is a way to nowhere, or to the curve-fitting, which we should avoid. Instead, you can apply some easier filter based just on prices.
Last, but not least, Japanese patterns are natural for us, human beings. By just merely looking on the chart, I can see it, feel it, without the necessity to calculate complex formulas. There's a psychology behind patterns, which is hard to be noticed by calculating a very complex formula.
The good software is very helpful in terms of being objective. It may help you to see what works, and what doesn't, to find your market edge. And for that reason I see Candlescanner software and website at least as a great learning tool for the beginners.