Spent most of today trying to solve a memory corruption problem which occurred when loading large models... Temporary solution working now.
Also spent a bit of time setting up better light attenuation and some debug drawing modes so I can see all the data being fed into the final lighting pass of the algorithm. Here are some screens of the various FBOs.
Overall I'm still a bit behind where I'd like to be, but making progress.
Position, diffuse, and specular, and normal maps.
Combined in the lighting phase via BRDF: