As I mentioned earlier, the barrels also move during recoil, generally as much rearward motion as any tilt-barrel auto. Ditto for the Lahti. The difference is that the barrel does not tilt. It is a simplified motion (at the expense of a more complex toggle-locking). On all Browning-based designs (Glock, SIG, Colt, others), the motion is in two planes, rearwards and tilting down (either in an axial rotation as in the Colt, or in a deflected vector as in Glock, SIG, or other, the result is the same). The tilting is really what causes potential accuracy issues, but there are many sources of accuracy degradation and tilting designs have proven accurate. In the Parabellum (or P38, Lahti, Beretta, Glisenti, Nambu, others), the barrel moves only in one plane (and then, essentially linear) and that is rearwards. There is no tilt. There is the second plane of movement still, but in those designs, the dropping or rising comes because the locking mechanism moves independently of the barrel. The toggle moves in the Parabellum, the tilting block in the Beretta/Walther moves.
Tilting designs have the barrel itself do the locking to the side. Since you have to have the vertical movement to unlock the pistol in all non-rotating recoil-based designs (at least, you have to have second-plane movement, a few rare designs have horizontal movement of the locking block or wedge and this ignores gas-operated designs completely), and since the lock-up is simplified, the barrel itself has to do the moving, in most cases at the rear.
As to whether or not a stamped sub-gun could have done the task, the mechanics of the operation are not lost on me. I know them quite well. Walther felt comfortable with blow-back (or, if it sounds better, unlocked recoil) designs, indeed, their P38 started as one. The bottom-feeders like Lorcin and Jennings, plus the step up Hi Point, have proven without a doubt the marketability of higher-power blow-back designs. I am not a fan of them, but there remains no real mechanical reason why Glock and company have not made them, too. Carry weight and overall balance are not mechanical problems but rather ergonomic problems.
Locked designs have advantages when it comes to service pistols, particularly in slide weight and size as well as ease of slide operations. However, consider the Ruger Mk II. It is a target pistol, not necessarily for competition, but for target shooting and perhaps hunting or trail carry.
Could one design a centerfire on the same lines for the purpose of target shooting? Absolutely. If using unlocked recoil/blow back, the rear slide could be made quite heavy and use dual-recoil springs like the Walther P38. This would place the balance of the pistol directly over the hand and all movement behind the hand. Weight of the pistol would not necessarily have to be excessive to accomplish this, and the design could be striker-fired like the Ruger or even hammer-fired like a Mauser. You could lock it by placing a sleeve around a locking bolt, not unlike the curved oprod in a VZ-52 rifle, and have it do the rearward movement instead of the barrel, leaving a locking post in the middle which serves as the breach-face. The actuator could then unlock the action when it reaches a certain point (the tilting lock would be ideal here) and the allow the entire bolt assembly to more rearward.
Not liking that, the HK method could be used in a similar way. Or, you could produce a modernized Nambu and get similar results. Remember, there were two different Nambu designs which were good pistols if you ignore the pathetic chambering.
It could be done. That nobody does it reflects the desires of the engineers as well as the prevailing aesthetic of centerfire pistols. Folks like them a bit blocky. Early on, many designers liked the slender barrel concept. Lahti, Parabellum, Glisenti, Nambu, and Walther designs reflect this. But the slide-enclosing barrel, either the Colt/Browning or the blocky SIG/Glock, aesthetic has become the norm these days.