Finally finished up my stable/neutral wind study.
If I hold the 925 wind, and surface to 925 depth and lapse rate as independent variables to the surface wind gust as the dependent I get a correlation of 77%. R squared is about 0.60, so 60% of the variance in gusts are explained by those 3 variables.
The regression equation wasn't too bad: gust = 31.9 + (0.5 * 925 wind) + (0.9 * lapse rate) - (0.05 * depth)
Ran it against an independent data set (2006-2008, study compiled from 2010-2019). I found that my equation was within 2 knots roughly a quarter of the time, 5 knots just under 50% of the time, and within 10 knots 84% of the time. I'm pretty happy with those results, especially for a first guess on the forecast.
Edit: All numbers for PWM (surface)/GYX (aloft)