Larry- If I may make a suggestion... (you may have already considered this) If you coat the lens for high transmission in green and make the band of transmitted wavelengths as small as possible, you can eliminate the need for an IR filter, as the lens itself would act as one (of course you may get a tiny bit of residual IR but it's really not a problem to begin with).
Also, aspherical may not be the way to go. The "raw" output from the crystals does not diverge nearly as much as the beam from a laser diode.
In order to achieve the results you see, the beam is expanded with a tiny concave lens placed directly above the crystals. This beam then passes into the final lens for collimation. In most cheap greens I have found that the "expander" lens is acrylic. If this lens is removed and a single element collimation lens is used you can get a small increase in power and a cleaner beam. If you plan on replacing both optics and getting decent divergence, you may have to use a 2 element lens to expand and then correct the beam.