I'm starting to wonder if the wind direction fluctuating wildly due to lack of damping in the vane might be the bigger problem with their forecast algorithm's accuracy?????
Wind direction apparently isn't a factor in their forecast method. A 3n1, for example, doesn't even detect wind direction.
This is just speculation, but I wonder if the problem may be due to multiple devices (e.g. bridges and USB consoles) being registered under one account and the algorithm draws data from the the wrong sources.
Another thing I wonder about is if it could result from an improper latitude/longitude setting. Once I saw a person who was complaining that the forecast indicated the wrong time of day in the forecast (e.g. showing moon/stars instead of sun, or something like that). The problem turned out to be their longitude setting. (Even though the time-zone was properly set, it was using longitude to determine forecast time-of-day.)
There are some quirky bugs in MBW, and some seem to be linked to particular accounts for some reason. About the only thing you can do is report them to Acurite and hope they can figure it out.