[justin_vanw]

So what about my initial point? How does a mass moving in the absence of other masses know that a geodesic for it is a euclidian straight path?

I would say that it doesn't, and that inertial motion in the absence of other masses is meaningless, and I think this is a key insight that allowed the development of GR in the first place (since any attempt to introduce absolute coordinate systems breaks causality because you can have multiple outcomes from the same boundary conditions, at least in the Hole Paradox).

I would say in addition that every other part of physics that I am aware of can be described locally, for example you can detect if you are in an magnetic field (at least in theory) even at the level of a single proton (since it has a magnetic moment). There is NO test you can do locally to detect that you moving inertially in the presence of a gravitational field. If you can't detect it locally, either there is a huge coincidence (in this case the coincidence is that inertial mass and gravitational mass are identical) or you can't expect to find any messenger particles, since why would you need to 'tell' a mass that it is in a gravitational field when it can't even detect it to begin with.

[lmm]

I would say you can't have anything that would have mass in the first place without a field, which in turn means you have space and time.