Georgina,
my goal is not to make you rolling over belly up. My goal is that you and the reader may understand the argument of inconsistency. I will now try to demonstrate this argument from another, more experimental perspective.
There are two options the source can send out particles (or particle pairs, considering your model):
1) The rotational axis of a particle (or the common axis of a particle pair; I do not speak here of the direction of spin rotation!) is always oriented the same relative to the laboratory frame, namely parallel to the magnetic field
2) The rotational axis of a particle (or the common axis of a particle pair; I do not speak here of the direction of spin rotation!) is randomly oriented relative to the laboratory frame
Let's first look at option 1).
By looking at option 1) we first look at the case that the magnet orientation is fixed - like in the original Stern-Gerlach experiment.
Let's label that experiment as "experiment 1".
In experiment 1 it seems one can choose between the usual explanation for how the open-mouth figure came about or your model. Despite the fact that you haven't made it explicit how the force or the combination of forces for option 1) leads to the open mouth figure, it is nonetheless clear that your model does not have the usual explanation for that figure (because that would mean your model would incorporate the same assumptions Quantum theory does make).
Now, due to the trigonometrical dependence for the amounts of classical forces that could be involved for option 1) (namely Lorentz and Faraday) - namely an hitherto unknown combination of these laws that somewhat lead for example to a particle being deflected a little bit upwards and a little bit to the left - the logical consequence would be that if one turns the magnet, say 23 degrees off from the orientation relative to the laboratory frame for witch the original Stern-Gerlach open mouth figure was found, then according to your model this figure must change dramatically, since the trajectories of the particles are angle-dependent. Since such a change of the figure isn't the case experimentally, option 1) has to be excluded as a possibility.
Additionally, option 1) implies something more strange, namely that whenever we make experiment 1, the particle's orientation of axis is always identical to that of the magnetic field lines. For very big distances between the source and the magnets, this would imply non-local adaptions for the source to the actual orientation of the magnet. Therefore possibility 1) has to be excluded - since your model aims to exclude such non-local adaptions.
Now looking at option 2).
By looking at option 2) we first look at the case that the magnet orientation is fixed - like in the original Stern-Gerlach experiment.
Here your model may assume that all the possible combinations of angles between the axis of a particle send out form the source and the magnetic field, when combined with different amounts of Lorentz and Faraday forces, results in the open mouth picture. That assumption can be characterized by saying that this combination of known laws together with the multitude of different particle axis' relative to the fixed axis of the magnet field is an hitherto non-analysed special case such that somehow those particles whose axis' are oriented at certain relative angles apart from the magnet's vertical axis (relative angles means the angles that the particle's axis establish with the vertical orientation of the magnet field) are those that must fly only a little bit upwards and a little bit left for example. In your model, this then would depend on the particle's initial orientation of rotation axis and its direction of rotation (CW or CCW) and maybe on some additional assumptions.
If one now turns the magnet, say 23 degrees off from the orientation relative to the laboratory frame for which the open mouth figure in experiment 1 was found, then in your model this figure should somehow be conserved under such a rotation.
If one assumes this, then consequently, the subtle combinatorics of Lorentz force, Faraday law and a multitude of different particle axis' orientations relative to the laboratory frame together with the discrete parameter of clockwise and anti-clockwise spinning direction (and perhaps some additional assumptions) seems to be responsible for the original Stern-Gerlach results.
The crucial point here is that one can easily experimentally test whether or not these subtle combinatorics with or without some additional assumptions are indeed responsible for the open-mouth figure.
All one has to do in such an extended Stern-Gerlach experiment is to filter all particles that go upwards into a channel and once again start a new Stern-Gerlach experiment with that subset of particles.
We now label this new experiment as "experiment 2".
Experiment 2 will not be done with a horizontal aperture (as has been done for experiment 1), but with a circular one. That alone would lead to a spot in the upper half of the detection screen. If that spot has established, we move the magnet a little bit to the left and wait until enough particles went through it. After that we move the magnet a little bit to the right side (in reference to its initial position before having moved it at all) and again wait until enough particles went through it. We do so systematically until we have sufficiently simulated the whole horizontal (and therefore also vertical) dimension of the original horizontal aperture.
Now, concerning your model, all the particles that came out of the first magnet in experiment 1) have an axis of rotation aligned parallel with those magnet they will go through for experiment 2 (the latter simply being a re-measurement of the subset I described). That subset is composed of particles that may have clockwise as well as anti-clockwise rotation of axis. And according to your model all these particles' orientations and spin directions are maintained when measured again with the same field orientation - these properties are not changed by an identical re-mesurement. That means that the resulting figure on the detection screen should be such that only the upper part of that picture is realizable, since all the particles have been prepared with the net force of "upwards" in the first measurement and re-measuring them with the same procedure that produced that net force will not alter that net force and thus will not alter the trajectories they will follow when re-measured.
Even for the - illogical - case that one claims that the subset of particles is somehow composed of only those particles that give the full open-mouth picture (say, for example only clockwise particles), this open-mouth picture will be distorted when the second experiment is repeated with the second magnet turned, say 23 degree away from its position it had in the first run of the second experiment.
Let's label this 23 degree experiment as "experiment 3".
Quantum mechanics now predicts something other for experiment 3, namely that such a collective re-measurement will result again in the full open-mouth figure - independent of what angle the second magnet has relative to the vertical laboratory frame.
Note the internal contradiction derived here in your model:
An identical re-measurement should neither alter the orientation of axis nor the direction of axis rotation. Hence this would lead to the result that all incoming particles again leave the upper channel and only the upper part of an (unknown) figure will appear on the measurement screen for the second experiment (or alternatively and illogically both parts, what makes no difference to the argument!).
Now note also that experiment 2 in your model is equivalent with assuming option 1) to be true. But we already figured out that option 1) is untenable for your model since it predicts that independent of the open-mouth figure being merely replicated in the upper half or replicated in the upper and lower half of the screen, this figure must change dramatically when experiment 3 is applied to our subset of particles.
So option 1) is untenable for your model to obtain the results of experiment 3 and the only possibility that an identical re-measurement as well as measurement of kind "experiment 3" produces the upper as well as the lower part of the open-mouth picture would be option 2). But for the these two re-measurements we already eliminated option 2) by selecting only particles whose axis' are parallel to the second magnet's field lines. Consequently, your model cannot account for the experimental fact that a full second experimental run of experiment 2, but this time with a changed magnet orientation of 23 degree (called "experiment 3"), will produce the same full open-mouth picture (it cannot account for this result because the deviations from the flight-paths of the particles in your model *depend* on the relative angle of the particles' axis with the orientation of the magnetic field lines and hence the resulting figure must change for experiment 3 - when compared to experiment 2). Thus, the resulting figure predicted by your model wouldn't be any more the quantum mechanically predicted and experimentally confirmed open-mouth picture.
In other words, if your model claims to reproduce all SG experimental results, it must claim to satisfy two mutually exclusive conditions for one and the same subset of particles, namely that each and every particle should be parallel to the magnetic field lines and at the same time being non-parallel to the magnetic field lines. Since this is logically impossible for a non-quantum explanation, your model is either equivalent with quantum mechanics (with all its weirdness) or is simply contradictory and hence false.
Note also that a third option, namely
3) The rotational axis of a particle (or the common axis of a particle pair; I do not speak here of the direction of spin rotation!) is EITHER oriented the same relative to the laboratory frame, namely parallel to the magnetic field OR orthogonal to the magnetic field
does produce the same contradiction I outlined since that option 3) does only encompass two out of the many different axis orientations relative to a given magnetic field direction the source can produce according to option 2).
I hope that with that analysis you better understand why many physicists are not convinced that the results of those particle experiments can be explained by the kind of attempt you gave. The fact that they aren't convinced is not due to some ignorance, but do to the laws of logic.
And I hope that what I wrote here makes clear that experiments with little gyroscopes are totally non-conclusive regarding the question whether or not your model is a feasible explanation for the experimental outcomes.
I additionally hope that this time I was able to more clearly show the internal inconsistency that I see in your model. If you have nonetheless questions, just ask and I will answer.
Concerning the "illogical" alternative to a local dynamics, namely non-locality, faster-than-light influences and non-existent particle properties until measured, I will write something about later when the dust of how one must define "consistency" consistently (instead of defining it inconsistently) has settled a bit more.
