Not to make (too) light of it, but I always got a chuckle when someone in the scammy forex world would refer to it as "grid trading." "Grid trading gives you the AIDS" (Gay Related Immunodeficiency Disease). Crass, I know.
Sounds like he is simply trading around a core position. This keeps you engaged with the prevailing trend with larger size while also taking smaller positions on the counter moves.
It can, of course, make perfect sense. You're out commissions and spread, but you've posted trades. If the value of posting those trades is larger then the cost of posting them (commissions and spread), then it makes sense. It's really no different from trading SPY vs IVV to max out your buying power and getting your broker to auto-cancel all your pending option orders without cancel fees or impacting your fill/cancel ratio. Or keeping $1 in your Oanda account to get access to the "real" (vs "game") data API. Some brokers have activity minimums to qualify for subsidised feeds, API's, or research. For those brokers, the ability to make breakeven trades (before commissions and spread) can be valuable. In the past (~12 or so years ago) some of the more reputable MT4 brokers "guaranteed" execution on stops to exit, but not stops to enter. Some events (like NFP) were exempt from this guarantee, but plenty were not. Going into these events with offsetting positions and exit stops at two of these brokers was, at the time, easy profits. In some markets, there are more and lengthier risk checks on orders "to open" than on orders "to close." Obviously, on offsetting positions, closing one side is opening a net position, giving you a slight speed advantage. In dealer makets, you sometimes find arbs. Usually it is the dealer making bad prices under some predictable circumstance. If you trade only at those times, they'll figure it out. Burying the arbs in numerous breakeven trades can delay the moment when the dealer figures out you're gaming him.
Yet it is not a stop loss, as there is no "loss". You are net flat. There is nothing to protect. If you want to buy a break out or sell a breakdown, you can set buy stop and sell stop order to open. There is no reason to have two accounts with margin requirements as a strategy. They ONLY reason to have offsetting positions is when you have to accounts, which is fine for those with material assets, and want to exit a trade but are signed into the other account. That is a trade of convenience not to make money.
I can trade both ways with my FX account, something i played with, in a maket turned against me, cant take a loss at a short enrty point, so take short trades, take profits off that and wait for the long to come back. Kinda works but that short becomes a loser, you take a profit eventually on your long then short stuck and always end up with 1 trade counter trend then cant always be around to compensate for that trade and a big move while not around kills you, or have to leave a counter trend trade in play while not there. So its valid in ways, but not without its issues.
This thread is like driving by a car wreck. You don't want to slow down and look but you can't help yourself.
Most traders have more than one account these days and generally everyone, who actually trades, does different things in them. The interviewee trades mainly CL using 10 second bars and some ES with the help of automation. Doesn't seem like ET has many sub one minute scalpers. Below are images of the same order algo running during same time in different directions. (PST) LONG SHORT
This describes a very basic, but fairly realistic IMHO straddle scalping example for QQQ. Old post (2003), but shows the basics in a simplified way: https://www.elitetrader.com/et/threads/gamma-scalping.15395/
The problem with any long or short straddle scalping strategy is that you are saying the market has mispriced ATM VOL. That certainly happens over a short period of time, but you need a process to predict future actual VOL. If futures price movements do not materially exceed (for long straddle) the VOL you bought, fees and decay will bleed you.