1) Gauge theory only includes Special Relativity, a quantum theory of gravity is simply an unsolved problem so far.
2) In GR gravity is a 'fictitious force', which would cause less problems if they had use the term 'apparent' or 'inertial' force. It is a force observed which is an artifact of your reference frame.
3) In natural units 1 unit of "time" is a huge dimension compared to the spacial dimensions and the Earth is following the geodesic or the "straightest" path around the sun due to curvature in that dimension.
4) While 'radially inward' does happen with inward falling, that is more of an issue about trying to extend euclidean space past it's useful domain, you are going non-local at that point. That radial inward path convergence is length contraction in the spacial domain, which is smaller than the more observable time effects but it is still an artifact of one's reference frame.
Most proposed quantum theories of gravity view it as a force field that is like the other force fields in QFT. Time will tell if they can accurately create a model that works in a way that can stay in a euclidean space.
Everyone has their favorite. To be honest I like geometry more than algebra so I prefer GR, but really I just someone comes up with a model that works. Noting my bias, I think eventually QM is going to have to give up on the quantum mechanical hope for gravitation and resort to a non-euclidean solution. As I can't imagine 4D I would be happy for someone to prove me wrong.