Designed to be long action,
https://photos.google.com/share/AF1QipMivUp-IOvnNTZKHA69mOUhyyt0cwtRTZ5SnjJXk350V0xgauDu6M2oVRpRspBCUA?key=ODEwa0xCYVNWVE1nQkoydlBVSFFhX2twUWZROVJn
So since the bolt face only needs to come back about 1/8" past the catch on rearward stroke (dead blow of the buffer will do the rest), pull the buffer and drop a quarter into the receiver extension behind the buffer. This will have the bolt face stopping about 1/8" in front of the back edge of the ejection port window, and plenty of room for the spent case to be cleanly pivoted off the bolt face and out the ejection window at rear dead blow stall of the buffer.
Now on the extractor off the bolt face, use a small jewelers file to remove any burs found on the side of the extractor claw, as well as the sides of the rim relief channel on the extractor as well.
Hence burs on the side of the rim relief channel will prevent the extractor center snapping all way down tight to the case rim itself.
Burs on the side of the extractor claw will prevent the center section of the extractor claw from being able to get a good bite on the rim during the pull.
Next, use a spent case to push the ejector into the bolt face off the extractor, and make sure that the ejector is not biding in the bolt face channel. Hence if ejector side slot was not lined up correctly when roll pin was pressed in, could be causing the extractor to bind up. If you do have binding, then pull the extractor, clean up the ejector channel and bolt face channel, and install new roll pin (while holding extractor in correct index so new roll pin and be punched in cleanly without bending it again.
If you still have problems, install a #60 O ring around the extractor spring, and put the extractor back in the bolt, The O ring will add more tension to the spring, to make dam sure that the extractor is not dropping the case during the pull out of the chamber, to the all the way back position.
Note, chamber will self polish via live fire and cleaning in time, so when the O ring only last about 500 rounds and you need to remove it, should not be needed again.
So long as bolt will lock back on last round fired out of the mag (catch in front of the bolt, and not just catching the bottom of the carrier), bolt unlock timing is fine.
As for chamber (Ogive leade space), free drop the round in question into the chamber with Bolt removed, use a wooden dowel to tap the back of the round in the chamber to make sure that case shoulder is tight to chamber shoulder, then tilt muzzle up, and live round should fall out of chamber. If it sticks and you have to use a cleaning rod to knock the live round back out, take a good look at the round. If marks on the case from say a the chamber not chamber brush cleaned with CLP from the start, or is the problem bullet embedding into rifling from the ogive of the bullet being too long for the ogive of the chamber. Also to point out, if the later, then spent primer is going to tell the story with is flattened out to the sides of the case primer pocket, since the bullet embedding at loading is causing higher than normal working pressures of the round.
Middle primer border line, while one of the right is showing working pressures over the top.
Simply, with the 168gr ammo, it uses slower burning powder, and this increases the barrel gas pressure to unlock the bolt a little earlier. With most rigs in 308 set up to run mil spec ammo, not a problem, but since it does mean that the bolt is going to unlock a littler earlier with the bore residual pressure a little higher, need to make dam sure that the extractor can pull the case out of the chamber, and not drop the case on the way back as well.
Yes, we can play the heaver buffer game to slow the unlock back down, but may cause problems with the rig not correctly stroking when feed lighter ammo instead. So as long as the above tweaks solves the problem (clean ejection of the spent case) leave the buffer in play for now. If still having problems like not enough dead blow stall for the spent case to pivoted off the bolt face with the above, then may have to go to heaver buffer to control the stroke/produce the correct rearward stall instead.