Relapse said..Doesn't cover Bom models but a really good overview of global weather models.
Good explanation of the various computer models. Which model is best depends a lot on the spot in my experience. On Cape Cod, which sticks out into the ocean, the higher resolution models usually under-predict the wind, since they over-estimate how much the wind drops close to land. On Cape Hatteras, the higher-resolution models seem to be more accurate, maybe because the land area is very narrow, and the gap to the main land is large.
All models I look at at a regular basis tend to underestimate thermals and local effects. At Kalmus on Cape Cod, we often get 10-15 mph more than the highest forecast if (a) the direction is WSW, (b) humidity is low (for Cape Cod), and (c) the difference between air and water temperature is small enough, or the predicted wind is strong (20+ knots). Sun also helps, although we sometimes get the WSW boost even on cloudy days. The model that often gets closest on WSW days is the IW-WRF model, which takes local historical data into account.
For late winter and spring in Corpus Christi, Texas, thermal forecasts are even easier: sunny and onshore winds = thermal boost, which can be 10 knots or more.
So if you have weather stations at your spot that measure the wind accurately, or get to play enough days on the water, check out the various model predictions and compare them to what's actually happening. With time, you can often learn which models tends to be best, and what factors (like wind direction, sun/clouds/humidity, temperatures, etc.) can be useful to modify the forecast.