The best range at this time with the technology is generally between 20 (because of ground clutter) and 120 miles from the radar location (because you start to get "range folding").
There's also the issue that the radar beam's height above the surface increases with distance from the radar, both due to the curvature of the Earth and because it's usually tilted at 0.5 degrees above the ground. So as you get farther from the radar site, the radar will start to only see the upper portions of storms, and at enough distance it might miss some low-topped storms entirely.
In addition to the much smaller population densities, one of the main reasons coverage is not great in a lot of the west is because of mountains, which block lower portions of the radar beam. It can get quite difficult to find a site where there will be unobstructed views in most or all directions - and the suitable sites that do exist are often too high to be all that useful. For example, the Cedar City, UT radar site is over 10,000 feet above sea level (more than 5000 feet above nearby valley floors) which makes that radar useless for low-level storm analysis.
I've actually talked to some of the meteorologists at the NWS office in Salt Lake City about this. They would love nothing more than another radar site somewhere northeast of Price, and sites near Delta and the Four Corners would also be quite useful. The main issue, as is usually the case in government, is $$$$$$. Radars cost a
lot, and the federal government isn't really all that interested in dropping a bunch of money on a radar that might improve coverage for a few thousand people in the middle of the desert. In fact, in order to get the current WSR-88D network funded in the first place, the NWS had to lead Congress into believing the radars would be more effective than they actually are.
