At least this should be true: measuring incident light at the shadow side of the subject (i.e.the opposite part of the subject lit by the main light) will place middle grey PARTS of the subject at the shadow side middle grey (given normal development and appropriate film speed). This is roughly definitional (and what Mark has in mind I suppose). If this is what You mean by "shadows will be on middle gray", then something is wrong with BTZS (probably not) or this understanding of the system, if it's supposted to be a proper guide to how to expose/develop for a proper picture. Normally, if the main part of the picture is in the main light with highligts and middle grey these part should show up as highligts and middle grey, middle grey parts in the shadow side (grey chart in the shawow!) should not (normally!) show up as middle grey, rather steps darker.

Well, someone more acquainted with the BTZS, SBR and EFS will probably give you more advice on this subject. (Personally my standard method is to measure the incident light on the shadow side and underexpose that roughly 2-3 steps, if the contrast is quite normal, and just 1-0 if I plan to do a more soft development due to the contrast, since in that case some shadow density will get lost. This is not very precise as I suppose BTZS pretend to - but works normally quite nice.)