Scale dependent group renderers would solve several rendering problems, especially for use in web maps via GisServer. The idea would be to allow multiple renderers to be assigned to each layer in ArcMap, with each renderer having its own minimum and maximum scale settings. In this way, a single layer could have one renderer when zoomed far out and a different one when zoomed in close. In fact, since each renderer in the group would be independent of the others, at some scales more than one renderer might be used to symbolize the layer. A good use case is where you want to render a polygon layer as a solid fill when zoomed out far, a hatch fill when zoomed in closer, and no fill when zoomed in close. Cuirrently, three separate layers are generally used each with a different renderer and scale limits. This is okay for just drawing, but especially in web maps, this technique has bad side effects. When a user identifies features by clicking the map at a certain scale, the selection is on one layer, and then when zooming in, that layer is no longer visible. Even worse is the case where a web map has a search tool that is meant to target a specific layer. With 3 different layers used to render, whcih layer do you search and select from? Selecting from all 3 would be badly redundant and confusing to the user. It would also require special code to coordinate a selection across multiple layers.