Any plans on optimizing Nanite Tessellation in the future?

Making an entire lansdcape or ground of nanite programmable rasterization doesnt look like a good idea to me in terms of performance. Have someone considered in for example having a tessellation disable distance like we already have with WPO?