given the following: the main OIII line is at about 500nm, and googling "pantone 500nm" yields this page:
http://encycolorpedia.com/00b8e0"In the RGB color model #00b8e0 is comprised of 0% red, 72.16% green and 87.84% blue."
i've usually blended in my OIII data to my green channel data at 72% strength and at 88% strength into blue. you can do this by using linear fit to normalize the signal strength of your OIII to the green channel, then use pixelmath to scale the OIII to 72%, then average the green and scaled OIII. repeat for blue channel on a fresh copy of the OIII.
i have no idea if this is a legitimate technique, but that's what i've done in the past.
rob