DOCTORSAIB, you are right.
Bernard-Soulier is a defect in platlet adhesion glycoprotein, the one that can bind to vWB, therefore BS and vWD both giive you a failed ristocetin assay. However, in BS, the ristocetin is not corrigible with addition of normal FVIII/vWB.
The variable thrombocytopenia has been hypothesized to be either an inherent decrease in platelet lifecycle or ineffective thrombopoiesis.
Clinically, BS shows heterogeneity; variants include an antibody-mediated (against the glycoprotein receptor), called pseudo-bernard soulier. And even a case associated with Gaucher's Dz, which presented with a BS-like picture, including abnormal ristocetin. It was found patient had intact GPIb receptor and no antibodies detected. This was termed pseudo-pseudo-Bernard-Soulier.
Glanzmann's dz is a defect in platlet GP IIb/IIIa, for aggregation. Variants have also included an antibody-mediated type.