I've always assumed that, to a first approximation, Vyse will vary with weight (and therefore Lift); so Vyse at Weight W1 can be estimated by modifying the Vyse at MAUW by the ratio of the SQRTs of W1 and MAUW. Obviously there will be a small form drag element that's not lift-related, but what else is significant?
Engine power variation with speed.
Your calculation will correctly assess the ratio of drag and power required. Thus for glide angle and sink rate minimization, it works. However, as soon as you start adding speed-dependent power, it shifts the minimum of the
excess power curve.