The discussions of profiting USD vs BTC are missing the point.
The real question is, does the EMA strategy beat buy-and-hold (i.e., will I have more or less converted BTC after using the bot)? The answer seems to be it depends on the conditions of the market.
During this recent insane rally, I don't see any EMA beating buy-and-hold. Another algorithmic strategy I'm using similarly performed worse than buy-and-hold in this period. I have had success trading half-manually with this bot entering the short position and then manually buying back BTC when I have some gains and I see a trend reversal. EMA has a lag in tracking the trend reversal and the market is moving very fast.
I've been using the bot with close supervision for about 2 weeks and I'm up about 8% in BTC. The largest gain was made by manually buying back during the flash crash two days ago. YMMV.
bottom line, in none of my testing across either butterbot or several competing bots not one of them performed better than simple buy and hold strategy. Obviously there are a near infinite # of strategies I didn't have the opportunity to test, but as of right now, buy and hold is looking like the ticket to maximum profitability imho.