I'd rather validate computational methods using a 2D cylinder (infinite length). It would be easier (than sphere) to get cross section of the flow pattern in its wake. I'd want to see flow agreement to confirm the proper regime. A steady separation bubble from tail eddies is the characteristic form of interest. Then fine tune simulation parameters until the Cd matches expectations. Hoerner, Hucho and others provide documentation for reference. Sorry I don't have links handy. Maybe someone else can and will share.
