A factor of 10 for values in an inertia matrix is not insignificant.
I think the 6 commonly used model files are probably close (someone should closely inspect the diffs to be sure), but the two less-used ones, thor and zeus, seem very different in some sections.
I've noticed some of the rover SDF files have very different mass and inertia definitions for some links.
For example: https://github.com/BCLab-UNM/SwarmBaseCode-ROS/blob/112118744de6d3fc12b3f29c0b1918dd0eb4075d/simulation/models/achilles/model.sdf#L7-L17
https://github.com/BCLab-UNM/SwarmBaseCode-ROS/blob/112118744de6d3fc12b3f29c0b1918dd0eb4075d/simulation/models/thor/model.sdf#L7-L17
A factor of 10 for values in an inertia matrix is not insignificant.
I think the 6 commonly used model files are probably close (someone should closely inspect the diffs to be sure), but the two less-used ones, thor and zeus, seem very different in some sections.