Do I have to be the one to say that this thread is WORTHLESS WITHOUT PICTURES? Actually, knowing Eric, I am sure he will put together a really awesome video.
I don't know. There is a problem with these diodes and the problem is the price. They are so expensive/watt @ at least double the cost of a P73 combined with that, the risk is that to achieve even this ratio you have to drive this $400 pretty hard. I don't think there is going to be a lot of community enthusiasm. I believe that DTR's prices reflect the cost from his source. If these were cheap then he would sell them cheep and he would sell thousands. The only niche I see @ this time is for awesome line generators or for over the top red "Laserscopes" with large beams and 10's of Watts of power.

Has anyone tried a prism pair after the diode but before collimating lens ?
I tried a single prism at a shallow angle and adjusted the cylinders for the angle deviation. It did not seem to be any better or worse than the cylinder alone for a given trade off in near field sizes far field divergence. This was a pretty poor test and a small improvement might not have shown up with such a conservative expansion.

I have an inspection microscope, but I think it would be useful if someone with an electron microscope or at least a high power microscope with measuring software could look at the facet. I suspect this is a closely packed, multiple emitter bar and if so then there are production, slow axis aspheric collimator arrays designed for specific emitter size and pitch that might be mounted in front of the existing FAC to improve the lateral divergence.