Well written. Simple and to the point. Could this idiosyncratic risk be captured in the contect of total beta and market beta as a whole? total beta = market beta / correlation with the market and market beta = covariance (R_m, R_s)/ Variance (R_m)