Thanks for the words of support guys!
When I would look at a semi auto rifle, I would always imagine how certain things would affect a bolt gun if they were handled the same way. The goal was to take what 'I' thought were problem areas, and work towards eliminating them, or greatly reducing their influence on shot repeatability.
The Timberwolf and Coyote for instance, have a C shaped cross section, and a very stiff bridge of steel connecting the front ring and rear ring. If you look at the cross section of a Remington 700, the right side action rail, is tiny. Barrels typically oscillate vertically, so an action weak vertically will flex and thus allow the overall whip of the barrel to have a higher amplitude. You can say that a 40X single shot action is more accurate than a 700 repeater because the action is a lot stiffer.....coupled with some extra bedding surface, which at the end of the day is stability. Solid bedding contributes to a consistent vibration pattern and shot to shot consistency. A Coyote repeater action is magnitudes stiffer than a 40X single shot.
There have been endless debates in the black rifle forum about accuracy of push pinned receiver sets. I used to have a Colt HBAR that I built with a DCM floater tube, Krieger heavy barrel and a Jewell trigger. Because the lower was a large pin, and my flat top was a small pin, I had to use an adaptor pin. By design of the adaptor, rotating it caused the large lobe to act like a cam and you could remove all the play by forcing the upper down into the lower. I didn't have an 'accu-wedge', but used a small square of wide rubber band dropped down into the rear pocket so I had to push the upper down to get the pin in. If I shot it without these preload steps, and just shot it loose, it's best groups were always worse than the worse groups when it was preloaded. Nobody can tell me that loosening off the action screws off a bolt action is going to have a positive effect on accuracy.
The upper of the Direwolf is bolted to the lower like a bolt action, there are locating bosses that centre everything up. Cleaning is achieved by removing one bolt from the buffer tube mount, and pulling the carrier out the rear. This buffer tube mount has lugs on it to locate it into the upper. By rear loading the carrier assembly, it allowed me to address one other issue seen in the AR10, but not an issue in the AR15. Carrier droop. On the AR10, the rear of the carrier is reduced in diameter to fit off the shelf AR15 buffer tubes. This leaves a fairly short set of contact rails on the carrier, with a lot of weight cantilevered out back. The buffer tube mount on the DW has a collar that fits into the large bore of the upper, and over the back end of the carrier, supporting it. While the AR10 is supported while in the buffer, it sits at rest and firing, at a droop. The massive bolt and carrier on the Direwolf operate with buttery smoothness. The bolt head itself is huge, really huge. There is a lot of shared design with the AR family in terms of operating principal, but areas beefed up for strength.
When I designed the Twolf, I increased lug contact 55% and shear area 210% over a Remington 700, which is available in .338 Lapua. The lug contact and shear area of the Direwolf is the same as a Timberwolf, over strong. After the CF adopted the Timberwolf, Colt Canada was tasked with a lot of testing of the system. Their goal was to determine risk to the shooter/spotter in a catastrophic failure. They started off with putting a projectile in the bore, 6" from the muzzle. This of course caused the barrel to burst, and the high speed video is interesting, it looks like an alligator flapping his jaws. The witness paper showed no debris hitting the operators. Next up they jammed a bullet into the throat and fired another round. The over pressure came back through the firing pin, extending it back to the point of spring coil bind and when it released, it broke the sear on the trigger. There was no lug setback or any other structural damage to the receiver. IIRC they repeated this test SEVEN times in order to fail the action, and they couldn't. They told us that those tests on a C7 would reduce it to a bunch of shrapnel and that they had never seen a rifle survive that test, let alone seven times. I believe the rifle was repaired and was used for additional testing. Is the Timberwolf strong, you bet your butt it is! Did the lug contact need to be 50% more or the shear area 210% more? No idea, I didn't have FEA ability at the time, just how much bigger can I make this without making it into a club. Since the lugs and barrel extension share Timberwolf like strength built into it, I have faith in holding such a hot round. The barrel extension and barrel nut are quite large, which was in order to increase barrel mounting rigidity.
Every other AR pattern rifle out there has the hand guard attached to the barrel nut, with the exception of the monolithic styles. Back in 2005, I made a prototype aluminum stock that used a short section to mount the action, and a piece of aluminum tubing bolted to it, that was easy. I learned a lot about solid modeling and ended up machining the Twolf stock out of a big chunk of billet, which exists the same way today. We can attach sections of rail to the side via holes, and it has various top rails to facilitate inline NV and thermal systems. This same NV 'top' is the one used on the US M40 system today, which everyone back then tried to copy but we won the USMC contract because they honoured our 'first' of design. Now, there are aluminum stocks everywhere and the common theme is Mlok which is open source. You can see in the CAD rendering that the lower extends forward, and the hand guard attaches to it. Nothing mounts to the barrel nut. What is unique is it is the lower receiver that mounts the tube, NOT the barrel nut. The barrel is 100% floated with the exception of the gas tube coming back. Through some magic there, and the desire to reduce accuracy harming vibration, there is a captive plastic bushing in the upper that eliminates metal to metal contact.
The .338 Lapua is no powder puff, it burns basically 4 times as much powder as .223, and a bit more than double a .308. There are substantial forces at play, all of them trying to rearrange the parts after each shot. Having mechanical shot to shot consistency is the goal which leads to having an accurate rifle. The Direwolf LOVES 300 grain Lapua Scenar factory ammo. I've shot a lot of groups that were down in the .5-.6" range. All of this with a MIL spec trigger from a LPC LOL....Because of the size of the bolt carrier, the bore centreline is a bit higher from the lower. Thus, getting a hammer to hit the pin with AR parts is a bit of a challenge. The prototypes were fine with this trigger, but production models will have a proper hammer and a decent trigger pull. My favorite trigger by far is the Trigger Tech, and don't be surprised if that is standard in a minor modified version. Other calibres will be explored, 300/338 Norma for sure. No desire to do a .300 Win Mag. The DW is massive, and putting .308 into it is kind of counter productive, there are existing platforms that are available. With that said, perhaps an upper/lower set with the true floated barrel system in .308 size would be fun to experiment with. We are currently working on a side cocker like an FAL style to get away from the handle bolted to the carrier. I'd also like to add a case deflector.