What is the origin of magnetoresistance in silver chalcogenides?

Silver chalcogenides (e.g., Ag2Te) with slightly altered stoichiometry exhibit an unusual magnetoresistance. It is large and linear in field for magnetic fields up to about 6 tesla and temperatures between 5 and 300 K. [See this 1997 Nature paper].
Several possible physical origins of the magnetoresistance have been proposed.

1. A PRL earlier this year proposes is the material is a topological insulator with gapless surface states described by a highly anisotropic Dirac cone.

2. In 1998 Abrikosov proposed the materials are gapless semiconductors with a linear spectrum, doped to a small carrier concentration, and that only one Landau level contributes to the conductivity.

3. Parish and Littlewood's proposal that the key physics is that of a strongly spatially inhomogeneous semiconductor which can be described by a random resistor network.

4. I also note that a band structure which produces a non-zero Berry curvature can produce a linear magnetoresistance, according to p. 1984 of this Rev. Mod. Phys. [However, I suspect this is a very small effect].

The challenge is to come up with experimental signatures which can distinguish between the four different theoretical proposals.