The NWS folks over in my area (Charleston) tend to blend models more than pick one individual model out. I work pretty closely with them being in TV weather - I tend to put a lot of focus on micro-climates as well, since we have several in the Appalachians. For example, models tend to usually overestimate snow in the lowlands in West Virginia, but an inverted trough with southern storms almost always 'warm wedges' us to where temperatures will be a couple of degrees warmer than model guidances - and thus little to no snow. The NWS does this quite a bit too as well in my region.
There's no doubt that the Euro is the better model, but I would be wary of using any one particular model for anything really. I look at as much data (both current and model) both at the surface and especially in the upper levels as possible, and I still use programs like Bufkit to get a better understanding of the atmosphere so that all the puzzle pieces fit when I'm trying to explain the forecast to the public on TV.