Has there been any attempts in dmd to generalize implicit conversions to allow implicit conversion of a head-const to tail-const range? Similar to what already happens for dynamic arrays (slices). Are there any apparent show-stoppers?