What's the old saying, that a little bit of knowledge is dangerous? So, we have the old statement that moving averages (at least for scalping) are useless, as they're a derivative (and lagging), and also that one must watch the DOM, which shows what went off at what price. Oh, and that Al Brooks scalps a 20 ema on a 5 minute chart.
Re: MA's - yes, they're a derivative, and seemingly, being so, they're even less reliable for scalping than with other methods/timeframes. But the answer is simple, particularly in our current world of historical data and charting: see if you can find a predictable-enough response to something to be able to trade it. In the case of MA's, despite being lagging and often of a longer-timeframe, they may be such that enough are watching them that you will get a predictable response. And particularly bec. the OP is speaking of scalping, enough to merit a trade. Ditto with Pivot Points, despite being maligned, if they have enough of the market aware of them, they may produce enough of a response to be able to scalp it.
As to the DOM: the DOM does not show what went off at what price; it's orders placed, but not yet filled. I can't believe that that statement is still hanging out there without a response. Substitute T&S for DOM.
Al Brooks: this was the initial allure of him and his methodology. That "all you need" is a laptop, with a single chart, with a 5-minute chart of the ES with a single 20 EMA.... Alas, what came to pass was that it was a LOT, and I say again, a LOT, LOT more complicated than that. Personally, I think the way he uses the 20 ema (as that was the OP's initial question) is simply a matter of being Fooled by Randomness (for the most part). He may be successful, but if so, it's because of the myriad of nuances to his trading that takes his method from simple to anything but. A LOT of subjectiivism is involved.
So, what do I suggest in this regard? As I said, look back on your own and find what has worked, and then forward test to confirm it. You just may find that MA's have value. I know of one trader who uses one on his charts, and the gist of his methodology is too play for the trend where price pulls back to that 2-3 times, and then prepare as it goes into a deeper pullback after that. At the same time, I know of another trader - quite aside from Al Brooks - who does a similar style of trading as Brooks. How? Because if you spend thousands of hours of screentime you can intuit things that can't be enumerated otherwise. So those subjective aspects of their trading is ingrained, yet hard (impossible?) to teach others. Okay, enough for now, I imagine we'd all agree