After reading one orginal ref. of Eng. Ref . I feel some inconsistency maybe exists between Eng.Ref. and original source.
In Page 152 (PDF 191), it interpolates the two sky luminance distributions, but in its reference, Perez (1990) interpolated the "geometrical factor ", which is the ratio of sky luminance distribution to Zenith luminance. In different sky types, Zenith luminance is different even if sky conditions are the same
I do not know how EPlus actually works in this case. if it follows the Engineering Ref. some errors may occur. But it probably follows the method of the reference.
Many other sky models are developed now. is it possible to update the sky luminance distribution model in EPlus?