Your system is coming right along. I like the first graph very much. The second graph, does it use the same timeframe as number one? Going back to Magna's words on timeframe determining noise...possibly a way for rethinking number two and for thinking more about stops and targets.
Rethinking on graph one...using a smoothed moving average...Kaufman's efficiency ratio doesn't work well on an intraday time frame...you need another filter that is widely adaptive...Donchian's moving average crossover works pretty well. It provides a histogram much as MACD would; but also, gives a graphical polarized filter so you can see immediately whether to be long or short based on the moving average. Although the picture is 100 ticks per bar I still used a short moving average of 30 and a long moving average of 172 to create the indicator. Maybe it will help.
Bruce