I decided that I was going to do something about the Brass-To-Face (BTF) issue with my Glocks and set about learning all I can from here, other gun fora and YouTube. Then I'd go to the range to see which of my Glocks had an ejection problem and which didn't. Finally, I'd try, one at a time, applying a fix to each troubled Glock as suited its particular problem.
Ideally, the ejection should consistently be to the 3-4 o'clock position. Anything else indicates problems. Even if it's not to the face, a problem is indicated if the ejection is not consistent.
The Glocks in question are two Gen 3 G19s, two Gen 3 G17s, two Polymer 80 full size (G17 compatible) and one Gen 2 G19 (with an LCI extractor and bearing.)
It seems to be widely understood that the ejection/BTF problem is one of two things: 1) The slide cycles too slowly to mash the empty casing against the ejector properly or 2) the extractor does not hold the empty case well enough to convey it to the ejector.
The former can be caused by weak ammo or a too-strong recoil spring assembly (RSA).
The latter is caused by a bad ejector, a too-weak ejector spring or, perhaps, the wrong spring loaded bearing. Again, this is all according to what I found on line.
Some people have had luck with a pricey aftermarket extractor. Others say they have had success changing the ejector in a Gen 3 Glock to a Gen 4 ejector or polishing the extractor or installing a new extractor spring. There was no universal fix. Some people could find no fix.
There are some who, despite the widespread reports of BTF, dismiss it as operator error: "You're not gripping the gun properly." I dismiss such claims. I've seen people on Youtube intentionally limp wristing a properly-ejecting Glock in an attempt to cause BTF. It doesn't work.
One fellow somewhere claimed that the extractor isn't even necessary on a Glock to eject fired shells and that the extractor is really only there for extracting unfired rounds when clearing the weapon. Fired shells, he claimed, would be extracted by the blowback. This was easy enough to test so I did and found it to be nonsense. The extractor is very much needed.
Someone reported seeing a BTF Glock in slo-mo video that showed the empty casing striking the top round in the magazine instead of the ejector.
Prior to going to the range to test each of my seven Glocks (all 9mm), I did a test on the bench at home. I removed the RSA and inserted a clean, resized, empty case into the chamber, taking care to ensure that the case slid into place behind the extractor. With all the Glocks, I could insert the empty case behind the extractor and it would be held in place.
I then reinstalled the slide (minus RSA) and, holding the gun in an upright firing position, pulled the slide back smartly to simulate firing the weapon. I watched where the empty case went. In most cases, with every Glock I have, the case ejected to the right as it should. Only if I gave the slide a less-than-quick pull did the case drop down through the empty magazine well. So, "in the shop," the extractor held the case long enough to bang it against the ejector and cause the case to exit the slide. I was encouraged.
At the range, I loaded one round into the chamber from the magazine, removed the magazine, then fired that round and noted the direction of ejection. In each weapon, I tried 115, 124 and 147 grain ammo. Only my Gen 2 G19 (with LCI extractor and proper bearing) ejected properly (3-4 o'clock direction). In the vast majority of other cases, there was no ejection at all; the empty case fell down through the magazine well.
This tells me that, had there been a magazine inserted, the case would have been bouncing off the top round in the magazine and not off the ejector. The fired cases never made it to the ejector.
On one or two of the weapons, I got decent ejection some of the time when using 124 or 147 grain rounds. I even got proper ejection once with a single 115 grain round. But the majority of all the rounds fired resulted in the expended casing going down the mag well.
From this, I have surmised that the extractor does not in fact hold the casing tightly enough to keep a grip when in actual recoil. That is, the tests on the bench, were I manually pulled the slide back to effect ejection, worked fine. There, the receiver was not in recoil; I held it fixed horizontally. But the case does not seem to hit the ejector at all when actually firing. The only reason it would not is that the extractor loses its grip on the fired case when actually firing.
Anyway, I'm thinking that changing the ejector would do no good when, by all appearances, the case never strikes the ejector. So my thrust at this time is looking into ways to have the extractor hold the case more tightly.
Some people have "changed" the extractor spring. It was not clear whether they replaced it with another stock spring or some aftermarket spring.
One fellow said he "filed" the extractor to make it reach further and give it a better grip.
Multiple people said that the Apex extractor ($60) worked for them. Others reported that it did not work for them.
I'm thinking that, if the extractor needs to grip more tightly, it might be possible to fix this with a Gen 2 (non-LCI) Spring Loaded Bearing which is about twice as long as the standard bearing on the external part that presses against the rear plate of a Glock slide. Using one of these would put more compression on the extractor/spring and might hold a case more tightly. Might.
I'm going to put my old Gen 2 non-LCI extractor and bearing into one of the Gen 3 weapons and see what that does.
Anyway, though the BTF issue has been beat to death, I thought I'd share my experience, FWIW.