You may have a couple of issues to consider here. One is a case of evidence of pedogenic processes and the other a case of sufficient soil fines (<2.0mm) to be considered ‘soil’.
The Soil Survey Manual discusses what it takes to make soil and the differences between B and C horizons. The simplest distinction is a B horizon must have evidence of pedogenic alteration related to the soil at the surface (not inherited from parent material or of a ‘geogenic’ origin). This alteration must be more than simple disintegration of the rock structure/cementation. This includes the processes related in soil formation including additions, losses, transformations, and translocations. In the setting you describe with saprolite, then enrichment with organic matter, leaching of soluble salts, formation of clay minerals or oxides, or translocation of clay serve as evidence of pedogenesis and the conversion of C horizon material to B horizon soil. As to the degree this must occur, that remains a judgment call (B vs. C) on part of the describer.
The other situation I have seen is the weathering of granite into smaller and smaller angular fragments with little or no soil fines. The material, though unconsolidated and relatively easy to dig, consisted of insufficient soil fines of <2mm to be called soil. There were some organic and clay coatings on the faces of the fragments. The layer was described as a C horizon. The Soil Survey Manual generalizes this amount to be about 90% or more rock fragments which would be described simply as ‘gravel’ or other terms used in lieu of texture.
