Erratic ejection patch,
Pull all the way back on the charging handle, and look where the face of the bolt stops compared to the back edge of the ejection port window.
The face of the bolt should stop 1/8" to 1/4" in front of the back of ejection port window edge. The allows for compression of the buffer bumper, and to insure that the spent brass is not tagging the back of ejection port window edge on the way out (brass should be tagging the very end of the upper receiver deflector at most, and not the end of window edge/flat edge at the start of the defector)..
If the bolt face is flush with, or slightly behind the ejection port window edge with the CH pulled all the way back,
Band-aid way to solve the problem is pull the buffer and recoil spring, and start dropping down quarters in the tube to act as end of tube shims to get the bolt face to stop in the correct position in front of the back edge of the ejection port window.
The correct fix through, remove the receiver extension, and face the front of the tube so it can be threaded into the receiver the wind or two more needed, and ending up with the butt stock correctly indexed, and the end of the tube just kissing the center post of the buffer retainer pin.
As for distances to face.remove material from the bottom of the tube, easy math.
Tube is threaded is 16 pitch, so every spin of the buffer tube will be the distance of .0625" in or out of the tube in the threads.
So let say that the tube needs to be spin in one more wind, and .03" off the center post of the buffer retaining pin now.
So it would be .0625" minus the .03, and you would face/remove .035" off the tube so it can be spun in another wind and just kiss the end of the buffer retaining pin and index the butt stock the same as it is now.
As for if you need to spin the tube in more than one wind to solve the problem, then this may create a problem with the top of the tube now protruding past the inside top of the receiver threads. Here, just mic the distance that the top of the tube needs to be set back and remove that amount of material.
As for removing material, I mill vice the tube so it top is facing out towards the front of the vise in the mill. This allows me to kiss off to remove material from the bottom of the tube (now at the back) to set the DRO to zero to start with, and if the top of the tube needs to be faced as well, easy enough too, and let the end mill make the needed round step edge from the top of tube to the end of the tube, by stopping the mill facing cut at the center line of the tube as I am moving the table towards me (all cuts made on the Y axis direction front to back, and just using the X axis for the needed depth of cuts).
Note, if you run into this problem on a fixed stock receiver extension, easy enough to solve, and it just the Lathe brought into play so you can face the stop collar back instead the needed distance (also comes in handy when there is too much of a gap between the end of the tube and the center post of the buffer retainer.
As for re-blackening the cut edge of the receiver extension, little acetone to clean all the oils off the fresh cut tube surface areas, then some birchwood casey aluminum black resolves that problem.
As for the failure to lock up, start with shotgunning the upper open, stick your finger in the back of the carrier, pull back and ride the B/C all the way forward to back to make sure that there is not binding. If needed, pull the B/C, remove the bolt, and use just the carrier with carrier key to check for binding. Here you are checking to make sure that the sides of the key not binding on the upper receiver and charging handle U slots, and that the gas tube is correctly indexed with the carrier key.
Note, with the upper coated, may have gotten some coating on the inside of the receiver, and this is causing the key to bind up on the sides of the upper receiver U slot.
Next, pull the ejector off the bolt and look at the color of the insert in the ejector spring. If it is black, then it should not have any O rings or D fender around the extractor spring when the extractor is reinstalled the the bolt. The Black insert is a marker that the extractor spring is already over tensioned from a standard spring, and when an O ring or D fender is added, the tension becomes too great to allow the extractor to cleanly slip over the rim of the case as loading.
Extractor reinstalled, take a good look at the face of the bolt, and the extractor to that circle ring.
The top left bolt has a good extractor fit, while the top right bolt extractor fit is a tad too far towards the center line of the bolt face.
Lastly, using the above photo, look at the bottom middle bolt and notice the rim releif cut of the extractor. The bottom channel on the extractor should flush, or slightly below the recess face of the bolt. This allows the rim to slip easy into the extractor releif channel as the extractor slips over the rim during load.
Lastly, grab a spent case, hook in on the extractor, and cam the ejector into the bolt a few times. Here, you are just checking to make sure that the ejector is not binding in the bolt face channel for it.