That addresses an issue that nobody has mentioned before -- that the slide itself is an integral part of the problem (due to the fact that metal will stretch and return to it's pre-stretch state if it's NOT stretched TOO far!)
Do ALL semi-auto slides demonstrate this same sort of accuracy-affecting behavior or is this a characteristic of the basic M&P 1.0 lockup design? It appears that the 2.0 version may incorporate design changes that make it less of an issue. (All that said, I've got an M&P Pro in 9mm and I've found the accuracy impressive, so I'm not criticizing M&P accuracy.)
You address PART of the question of "why" accuracy can be affected in your following comments.
That you can "test this by pulling the slide a few thousands of an inch" and wiggle the muzzle doesn't seem entirely convincing -- since the forces being applied to the barrel, slide, and recoil spring when the bullet is moving down the barrel under intense pressure would seem to make that whole assembly more rigid and less easily moved in a random manner than in your "test" example. I understand, however, that the "test" shows that the potential for movement exists. That a bullet spinning down the barrel may apply forces most of us have have never considered, and such subtle changes in the bullet loads can also affect accuracy is also interesting.
To this non-gunsmith layman, that might suggest ways to offset these "random" barrel behaviors could include barrel bushings, a different lockup design or, as you mention in passing, tighter vertical slide/frame clearances.
I appreciate, however, that when you do all of that, you've begun to build a different gun with what may be different design objectives -- maybe trying to create a target pistol and not a service pistol -- designed for a different role, different reliability objectives, and much different production costs.
Some very subtle, almost invisible M&P 2.0 changes are probably a far more realistic approach.
Thanks for your informative response -- now I've got more to think about!
Thanks for keeping me thinking too, Walt.
Before I started making parts, I was a starving pistolsmith, so I had the good fortune to work on your everyday run of the mill 1911s, Glocks, XDs, Browning HPs, Sigs etc. Most of them have more robust slides, particularly at the ejection port region where the stretching phenomenon is most likely to occur. I believe the added material reduces the effect.
The M2.0 has considerably more material on the slides, and if you look on both interior walls just in front of the extractor and the vertical wall on the side opposite, you will see a noticeable step. While this is present on the 1.0, the M2.0 has about .008-.009" more material protruding inward towards the chamber sidewalls of the barrel. This does two things:
First, it adds thickness to the areas where the stretching occurs and second it increases torque lock of the barrel to the slide as the bullet rotates its way down the barrel. This coupled with the 1:10" twist rate greatly improves accuracy potential in the 9mm.
"That you can "test this by pulling the slide a few thousands of an inch" and wiggle the muzzle doesn't seem entirely convincing -- since the forces being applied to the barrel, slide, and recoil spring when the bullet is moving down the barrel under intense pressure would seem to make that whole assembly more rigid and less easily moved in a random manner than in your "test" example."
Walt, the force being applied only helps to lock the barrel and slide along the axis of the bore. This means that the only contact point being stabilized is the top locking ledge of the barrel against the mating surface at the front of the ejection port. When the slide stretches, vertical forces acting on the barrel become variable. I believe it has a greater affect on the M&P pistol due to the way they designed the system. Other pistols like Sigs, 1911s Browings and my favorite 3rd Gen Smiths use a mechanical means to apply vertical force to bottom lug of the barrel. The M&P simply does not. Given that it is a service pistol, perhaps it isn't really a big concern for most.
Our barrel is different because we have the bottom lug of the barrel mate to the locking block- this is what provides the consistent vertical support as the slide stretches. It is not new magic or snake oil. This is the way the Sig P series TDA pistols have done it for decades, and custom built 1911s usually have the bottom lug cut so that the barrel is not only vertically supported through the dwell phase but beds the barrel's top lugs against the slide lug recesses. In all cases, for optimal accuracy the barrel must be mechanically fixed relative to the slide until the bullet leaves the bore. For the M&P, this requires the barrel to be supported by the frame locking block.