I use m/a's to the exclusion of any other "indicator" in my trading. In my experience, using the faster m/a to identify entry and also as your stop, and using your slower m/a to confirm direction it is possible to find low-risk entry-points for high-probabilty trades. Look at 5 min charts for stocks like KLAC, SNDK and AAPL and observe the price-action using m/a's (eg 9.5 period + 34 period SMA's) together with "hammers", "shooting-stars" "and/or wedge patterns at round numbers, 30, 50 and 70 cents.
If both m/a's are lined up beneath or above price action, and the current movement of the stock is in the same direction, you have empirical evidence of trend. If they don't or are flat - esp cutting through the current price action - then you know to stay on the sidelines.
IMO based on many months of wasted effort, if you stick with m/a crossovers, you will get sucked into trying to find supporting evidence from other indicators. Trouble is, the indicator you select will almost invariably be based on the same data source as the m/a's themselves - so in fact it doesn't help. It just adds to the confusion. The key to all this is to find LOW-RISK ENTRY POINTS. If you are wrong - so be it and move on. So long as you aim for gains of 3 x your risk per trade you shouldn't go broke. Gains of 1 x risk or less is a sure way to the poorhouse.
I am replying to you as I remember going through the exact same thought-processes some time back.
Good luck,
raider1