Actually, HK416 and their Magazine were both in production at the time.
The HK magazines were in production because they are a handoff from the SA80A2 program - although whether that was a benefit to the 416 is a fair question considering that TACOM and the Navy have both cancelled use of the HK magazine.
As for HK being "in production", my understanding is that neither FN nor HK had an actual production line set up. They were producing a limited number of weapons for SOCOM and test samples; but neither one was capable of fufilling large orders for rifles at the time of Dust Test III. Of course, that is just based on what I've read and there is certainly a lot of misinformation out there. If you have a reliable source you can share, I'd love to hear more.
If the port is in the wrong position, is the wrong size, or something is wrong with the gas tube, those are changes that you THINK they would have tried already with 20 different AR-15 manufacturers. I, for one, think they got it right for a 14.5" barrel.
Several different manufacturers have tried different positions than the standard M4 carbine and with good success; but those models don't get the kind of testing that the M4 does. Most of the upgrades in the M4 (heavier buffer, beefed up extractor, feed ramps, better mags) are really there to solve the problems inherent in a shorter gas system using parts designed around the pressures in a 20" rifle.
If you read into the dust tests, they actually break down how many failures were magazine related. For the M4, that is 239 magazine related failures. Assuming that the HK magazine is 100% reliable (It's not), then that would still mean the M4 got stompped.
Yes; but the HK had 9 magazine stoppages, the FN had 19 and the XM8 had 18. I think we can both agree that a difference between 239 and 9-19 indicates a bit of a problem that has nothing to do with the rifle.
Reduce the magazine stoppages to levels typical for the other rifles and we are already 1/3 of the way towards a DI rifle that performs as well in that test as the most modern gas piston designs - even before we do crazy stuff like say, building a DI rifle designed around a 14.5" system instead of making do with a 20" system cut down to 14.5".
I don't believe that the stress on the barrel extension is any greater nor is the bolt being stressed any more. In fact, I would say that the carbine barrel extension would be subjected to roughly the same forces.
The barrel extension isn't under any greater force; but if you change the design of the bolt, you must also change the design of the barrel extension. The bolt IS under much greater forces, check out the Rifle Forum Reading Library discussion of midlengths for more details and links to the technical data. The problem is that the bolt in use was built for a 20" rifle - and like many well-engineered items, it is designed to withstand stresses far greater than what it would commonly encounter in a 20" rifle. However, with a carbine, the gas reaches the bolt faster (shorter travel distance + higher port pressures) and unlocks the bolt earlier in the cycle than with a 20". This puts much higher stresses on the bolt and reduces the operational life span of the bolt.
Even a simple thing like radiusing where the bolt lugs meet the bolt and redesigning the extractor would be a big benefit; but logistically, you have to support two different kinds of barrel extensions and bolts, so the Army hasn't warmed up to that solution. Take a look at the early 2001 SOPMOD Improved Bolt project or the KAC E3 bolt (or any of the other bolts developed in response to that) and you'll see some of the problems we have created by using a bolt designed for a 20" rifle in everything from 14.5" carbines to 10.5" CQB type M16s.