The ammunition I was using was 55 grain .223. I was using two types of magazines: a steel magazine from Black Hawk, and a Magpul polymer magazine.
In addition to the foregoing, I had the following clues:
|Bolt Carrier Group--Left Side|
|Bolt Carrier Group -- Right Side|
Further researching the issue, however, I found an article from Guns Magazine on cycling issues in ARs that noted that an action that is cycling too fast can mimic short stroking. An article at Defense Review indicated that this can result from using too light of a buffer.
There are various weights of buffers used in AR systems. The M16 rifle uses a 5.2 ounce buffer (and also a longer buffer spring--please note, though, that neither the rifle buffer nor rifle length buffer spring can be used in a carbine tube because they are too long). The standard buffer in the carbine is only 3.0 ounces. Colt also offers heavier buffers for the carbine: the H (3.8 ounces), H2 (4.7 ounces), and H3 (5.6 ounces). The weighting is because the standard carbine buffer contains three steel weights. In an H buffer, one of these weights is replaced with a heavier tungsten weight; the H2 has two tungsten weights and one steel weight; and the H3 uses three tungsten weights. I also came across information indicating that Canada's AR military rifle uses a 20 inch barrel but a carbine buffer-tube/stock, and they use an H2 buffer.
In analyzing my problem, I considered the possibility of a problem with the gas system. I saw no evidence of a gas leak around the gas key. I thought that a misaligned gas block was unlikely since the assembly was by a reputable manufacturer. A magazine issue also seemed unlikely since I had used two different magazines from two quality manufacturers, and was having the same problems with both magazines. Also, I was intrigued by the dings on the buffer from it striking the retaining pin at the mouth of the buffer tube, which suggested that the buffer was moving with too high of velocity. Finally, my buffer was a standard 3.0 ounces carbine buffer (confirmed by weighing on a scale) and, therefore, over 2 ounces lighter than what a rifle length system was designed to use. However, because I had not taken out a variety of ammunition, I could not rule out an ammunition issue.
Using a thoroughly cleaned and lubricated weapon, I headed out to the range again. I purchased a heavier buffer (4.4 ounces according to my scale, so roughly the weight of an H2 buffer). I also took out more 55 grain ammunition (Black Hills), 62 grain "green-tip" manufactured by IMI, and some 70 grain "match" ammunition. Prior to replacing the buffer, I tested all three different types of ammunition, with the same problems. I replaced the buffer and, although I had occasional issues with the 55 grain ammunition, the performance was overall much better. By the end of another 100 rounds or so, it seemed to be feeding reliably with all three types of ammunition.
In short, I believe the issue was a combination of breaking in a new upper and bolt carrier group, and using the heavier buffer, with the latter being the more important point. Although it seems counter intuitive, since the rifle gas system has a much lower operating pressure than the carbine system, the rifle gas system apparently accelerated the BCG to a higher velocity than a carbine length system would have done. Using a heavier buffer, I was able to mitigate this to some extent. Since I am still well below the weight of the standard rifle buffer, it is possible I might have achieved better performance with a slightly heavier buffer. So, something to consider if you are moving to using a longer rifle-length gas system on a lower set up for a carbine.