I read some information last night in a Fuji patent from 1987 where they made a high iodide core (up to 45 mol %) and a low iodide shell (about 5 mol%) with the aim being an average iodide of 7%. The theory they gave in this patent was that "bending" (I think they meant crystal dislocations) at the core/shell interface would contribute to high sensitivity because of the high iodide and the low iodide shell would provide better development. Then "holes" created as photons knocked electrons into the conduction band would migrate into the high iodide core leaving the electrons in the better developing, low iodide shell to reduce Ag ions. I guess that by having the holes move into the core that recombination was discouraged which would contribute to sensitivity.
Light absorption increases in the order of AgCl, AgBr, AgI but development activity increases in the order AgI, AgBr, AgCl.
The Fuji patent is US 4,668,614. They seemed to feel that this was all a novel idea so maybe this was one of the 1st core/shell patents. Didn't really discuss prior art other than research work by Klien and Moisar.
Later work talked more about graded and multiple shells.
-- Jason