In this representation, gravitation is not a separated force, but is instead simply the proportion of all collapse events that can be attributed to bosons from the massive body. These events are always directed towards that massive body.
Where two bosons collapse into a fermion, the position of that fermion will most likely be located in a direct line between the two sources of the bosons, because that is the first solution that satisfies the quantization condition. Where one of the bosons contributes significantly to the mass of a conserved particle, the particle will be displaced in that direction.
Next we'll consider the direction of the boson sources, relative to the conserved particle.
For the vacuum, bosons come from all directions, so over repeated instances of the particle collapsing due to an omnidirectional vacuum, the average displacement tends to zero.
When a massive body is nearby, that body will have collapsed and re-radiated many bosons from the vacuum, and although they share the same mechanism as vacuum bosons, they are directional, and 'compete' with the vacuum's bosons to collapse the test particle.
Where a significant number of these bosons come from a massive body, each such collapse will nudge the particle slightly towards the body. Repeated displacements become significant, and the particle drifts towards the source of the bosons.
The structure of a conserved composite body is generally 'elastic', and distorts when displacements are applied. This changes the individial collapse events that make up the sequential re-constitution of the composite, conserving the momentum.
Larger composites will need more displacements to achieve velocity, which helps us describe classical inertia and momentum in quantum terms.
While a boson propagates, its mass-energy acts as a phase operator on the waves of other bosons. Consequently, where one boson proceeds to overlap another, its mass-energy widens the phase window for its own collapse, and the collapse of overlapping bosons (a process analogous to the Higgs mechanism).
When these quantum fluctuations are integrated over classical limits, this creates gravitational force as expected, and even permits bosons of zero mass to collapse if massive bosons overlap them.
Vacuum energy will interact with a large body (fig.1: C), and radiate from it as bosons, again as vacuum energy. The more mass-energy body C has, the more vacuum energy it will collapse and re-emit.
As body C’s bosons radiate, some will collapse. With increasing radius, their area for interaction increases, giving a higher probability of collapse from vacuum energy, as per eq.4. Some of the bosons available to test particle A will be environmental vacuum energy, and some will have been emitted by body C. Where bosons from particle C are preferred, this results in a gravitational deflection (or ‘force’). The resulting approximation of gravitational deflection  is comparable to classical formulations.
[equation to be inserted]
The mean deflection ∇, which is independent of the mass of the test particle (fig.1: A), is the probability that the test particle will interact with the body’s flux p_b, rather than with the environmental vacuum flux p_v, scaled by the mean expected vector ∇_b between particle events where the particle interacts with the body’s bosons.
As for the direction of the interaction, the first opportunity for gravitational interaction is the point directly between the bodies (see arrows meeting, fig.1). Vacuum interactions will be uniformly distributed in all directions, unless there are some very large structures that generate flux. The direction resulting from its own bosons collapsing will depend on its structure, including any structural changes attributed to its momentum.
Gravitational waves are just the same as regular gravity, manifesting as a directional preference for regular matter to be deflected during its usual reconstitution process, using the flux of gravitation. However, a gravity wave, believed to have origins in extreme distant events, is a temporary variance in the flux density.
When this flux traverses across a large body, this may locally compress and expand matter within a relatively short length, possibly affecting the distance between mean fermion positions within the body, and its effects should be classically measurable. (a) If we assume flat space, as per our red-shift paper (2014), then we would observe red and blue-shifting of surrounding matter, for the period we are affected by gravitational waves. We could pinpoint the source from the (spatial) directional distribution of spectral shift; (b) lengths within rigid structures would fluctuate.
Measuring it is a statistical calculation of the usual deflections, rather than a straight reading from a field. i.e. you cannot measure it more and more precisely; you'd just find fewer (or no) impulses within a smaller sample length, with regular gravitational interactions as background noise. Thus gravitational waves exhibit their own uncertainty relation, between the imparted momentum and the size of the sample.
We have reservations about whether a unified field theory can include gravity. Instead we suggest that, although we get gravitation ‘gratis’ with our mechanism, it might not be helpful to search for a unified field in the conventional manner. If our hypothesis holds, then gravity would be a fictitious field, prone to accountability problems.
The answer might instead lie in our approach, of finding a fundamental mechanism, with uniformly-defined entities, the simplest algebraic abstraction, and very simple rules. From that, we derive the conventional fields statistically from our understanding of their expression from fundamental structures, and with knowledge of the information lost when building such approximations.
In fig.2, all bosons overlapping the test particle have the same structure. Those partitioned as C and E are attributed to the gravitational flux from a nearby massive body. A are confined bosons, and B and D are bosons from anonymous vacuum energy. Fermions 1 and 2 are assumed to be an example sequence of fermion events within the test particle: respectively a virtual vacuum interaction, and (e.g.) a quark. 'Charge' is the property of conducting flux through the particle.