John....as someone who can generate a perspective with a tee square and set-square...and has done so professionally since 1972 I know 'all' about perspective, sciagraphy and reflection planes, so when I say the reflections look fine I mean they can be demonstrated/explained/reproduced as accurate by mathematical argument - assuming viewpoint.
In other words they are fine....
As for 'opaque'...are we presuming indices of refraction are what.....only acceptable if the conform to one person's criterium?
Perhaps what you mean is...it'd be nice if people could mess with how any such reflections are 'handled'.
If they are not to be adjustable that still will not make them 'wrong'...