I think the trick is in the interpretation. Never believe a forecast that tells you exactly what is happening exactly where and exactly when. I think the models get the general picture pretty well right, but often err with the exact timing or location. So the way I use weather forecasts is to look at the broad picture over a wider area and longer length of time, then allow for certain events to be slightly earlier or later than forecast or missing us entirely because a particular feature went slightly to the North or West or whatever. Does that make any sense?
I use the Met Office site but it's hilarious - I'll check it at Midnight and see that we're going to be applying Factor 50, when I get up seven hours later - it's raining stair-rods and sure enough, the site has updated to reflect this. Superb stuff.
I've swtiched to using the lightbox on the Tabard Tower now...
But I find the rainfall animation on the met office website very useful (and more useful than it used to be now that it shows shots every 30 minutes). Tracking the rain over the past few hours gives you a very good idea whether the next couple of hours are going to be wet or dry should that have a big impact on your plans.
I was based at the Met Office in Bracknell in the early days that that animation was developed, and I used to use it to decide when was a good time to cycle home. It won't surprise you to know that I often got wet.