I'd agree that local knowledge is valuable.
Weather prediction is a crap shoot, no matter who does it. And often it's over a broad area and you could be boating even 10 miles from the forecast area and be much different.
So many factors....
However, there's ways to somewhat predict the likelihood of nasty stuff, and one can do a bit of their own predicting. Need to start with actual data, weather stations, bouys, etc., and know where the weather station is. There are a lot of private weather stations that are poor, not mounted properly and don't give good data. Airports, bouys, stations on bridges without obstructions are usually excellent. Also, the radio sound info from the weather balloons is handy, launched twice daily, but can be hard to interpret.
Know the weather patterns... highs, lows, troughs, pressure, temp, wind patterns (18000 ft is a good indication of flows). Know the local situations.... hills, valleys, seas, lakes.
Often making last minute changes to plans is necessary because the forecast didn't come thru. It's pretty hard to predict accurate winds (and weather) much more than a day out, and even that changes.
And, there's private weather services, that sometimes do a better job than government ones, but ya pay for them.