Mesh size does not alter the melting behavior.  I question your source for the Silica melting point-- quartz undergoes a series of crystallographic transitions, one at 1064 F, one at 1598 F, and another at 2678 F.  These are all in the solid phase, not a solid to liquid phase transition.  The number you were given may be a suggested upper use temperature that is dependent on impurities, bonding resins, etc.

I don't have enough time right now to go into more detail, but other aggregate molding materials are used for sand casting, namely ziron (zirconium silicate), olivine, and various aluminum silicates.  You may want to investigate these further.  Also consult with a non-ferrous foundry on what they use-- you can find more info at www.afsinc.org.

We had tried pouring Semi Red brass(RG7 in European parlance)with nature sand of AFS 130 but the heat extraction from the sand was a problem.We could successfully cast our products with Shell moulds.
Silica sand of AFS 100 was resin coated and shell moulded for the purpose.

Foundry practice is as much an art as a science. Natural sand, clay as a binder and water go into the molding sand. The mesh size influences the surface finish, but finer sizes can sacrifice the internal soundness of the casting.
The sand doesn't melt; most heat extraction comes from vaporizing the water in the molding sand.
 Now you get to design a gate and riser system. Have fun!