Regarding the beam artifacts:
The diodes with the cleanest beam profiles I've seen are low power red (660-635nm) diodes.
Everything else seems to have some kind of artifact. The high power 200-300mW LPC 655nm diodes have a line that crosses through the beam and the ~400nm Blu-Ray drive diodes (PHR, BDR, etc) and all the ~455nm diodes I've used have a rectangular artifact similar to these green diodes.
This is an artifact caused by the emitter, I think. Removing the can from the diode does not remove the artifact and lenses only obscure the artifact to varying degrees.
High transmission optics like the G-2 lenses produce more severe artifacts because they are positioned close to the emitter to capture most of the diode's output. The single element design minimally deforms the beam profile so the artifacts are usually more pronounced with these optics.
The 3-element lenses are positioned farther from the emitter so more light is lost (hence the lower output power), essentially clipping off the fringes of the beam profile. The result from these 3 element lenses is a rounder beam up close but it's not efficiently reshaping the beam, just clipping it with a narrow entry aperture, so you get a terrible diffraction pattern at a distance.
The short version: this artifact is an inherent quality of the diode. Changing the lens or removing the can will not remove it.
You can use a narrow aperture after the lens to cut off the ugliness but unless the aperture is positioned at least a few centimeters from the lens, you won't notice a big difference.
DPSS and gas lasers produce much cleaner beams with beautiful gaussian profiles for some reason that has to do with the lasing medium. Don't ask me why that is - I'm a biologist, not a physicist.