Hello nopaixx,
transformation sub-sections for cameras contain 4x4 matrix that defined the camera rotation (3x3 matrix) and location in the internal coordinate system space.
If you apply 4x4 transformation matrix from the right from chunk section to these matrices, you'll get the coordinates in the geocentric system, so if you wish to get the coordinates in WGS84 system you need also to apply the coordinate system information (to project) to the product of matrices multiplication.
Probably, the easier way for your needs could be cameras export in omega, phi, kappa format or even just saving the estimated coordinates of the cameras from the Reference pane.