Quoted:
Before I start throwing around a bunch of stuff that sounds authoritative - I'm no SME, I'm just a mechanical engineer who has been trying to suss out the variables in my free time. Most of this is theoretical, and I'd defer to an experienced gunsmith or armorer.
I think the gas in the face phenomenon is a timing issue for ARs. To understand this, let's think about pressure at three locations: the chamber, the gas port, and the muzzle (bare or inside the suppressor).
Let's start with the unsuppressed case. After the bullet passes the gas block, but before it passes the muzzle, the pressure at the chamber and the gas port are similar. The muzzle is at atmospheric pressure
The gas system drives the bolt back to cycle it, but the bcg has not made any appreciable movement rearwards by the time the bullet has left the muzzle. Important to note, the muzzle is still at atmospheric pressure (more or less) as the gas exits, meaning there is very little resistance to flow and the pressure drops quickly. Gas is currently flowing out of both the gas tube into the receiver and out of the muzzle after the bullet exits. This allows much of the pressure at the chamber and at the gas port to dissipate before either are opened and are a path for gasses to travel. The residual pressure vents out of all three openings as the bcg cycles and chambers a new round.
Now the suppressed case. The difference occurs after the bullet leaves the muzzle and gas begins to fill the suppressor. This means the pressure at the muzzle is not atmospheric pressure, but is the "backpressure" of the suppressor and this "backpressure" decays slowly relative to the pressure decay in the previous case. This backpressure inhibits flow from the chamber and keeps pressures in the entire system elevated after the bullet has left the muzzle. The pressures are still elevated as the BCG moves rearward and begins to extract the spent round, resulting in both increased friction in the chamber (due to the brass being under pressure) and a sudden opening of a "path of least resistance" causes trapped gasses to flow to the chamber (which is exhausting at atmospheric pressure now, not the elevated pressures in the suppressor.) It wouldn't surprise me if in some circumstances gasses trapped inside the suppressor actually exit through the chamber and not through the muzzle of the suppressor.
The peak pressures in both systems are the same, it is the decay in pressure that differs. So, to minimize gas blowback (and action noise, and improve reliability) the key factor is to reduce the chamber pressure at extraction. There's three main ways to do this.
Firstly, if you extend the barrel and gas system, the extra barrel length already decreases the pressure. I attached a graph below from this excellent report on that subject:
Barrel length vs pressure The lower pressure starts you off in a better spot regardless of other mitigations. Of course, if your goal is to build a SBR, me telling you to use a longer barrel doesn't help anything.
https://www.AR15.Com/media/mediaFiles/132840/borepressure_PNG-1288635.JPG
Secondly, you can change the timing relationships. The idea here being to reduce the time between the bullet passing the gas port and the bullet exiting the barrel (dwell time) while delaying the BCG moving backwards to begin extraction. Altering dwell time has a lot of potential to interfere with proper function of the rifle while unsuppressed. It's a much more complicated problem to optimize a rifle for performance both with and without a suppressor. If you're happy with suppressed only, it's worth looking at 16"/rifle length, 11.5"/midlength type options. Beyond that, you want to slow down extraction. This is where increasing BCG/buffer weight and reducing gas flow (smaller port, adjustable gas block). One very interesting product for that is the LMT enhanced BCG. I noticed a couple people above said they'd had good luck with them (I'm still waiting on a backorder...we'll see if they beat the ATF or not). They changed the cam path for the BCG, which delays unlocking.
Finally, you can change the backpressure of the suppressor. This is the approach the OSS suppressors take and while I've never tried their products, it's an interesting approach. Beyond that, shorter fatter cans will be better for backpressure and worse for noise levels.
Personally, I've been toying with the idea of a pin and weld to 16" cut down of a rifle length gas barrel, coupled with a VLTOR a5 buffer setup and an LMT eBCG to try to make a suppressor centric AR.